NASA - James Webb Space Telescope - Discussion Thread 2

[LouScheffer]

Should Webb telescope’s data be open to all?

[...] But some astronomers question the practice, arguing that data from federally funded projects should be free for all to use. NASA, Webb’s primary backer, is facing an open data push from the White House and may soon end the restriction. Having so much Webb data locked away “doesn’t pass the smell test. It’s just not right,” says astronomer Garth Illingworth of the University of California, Santa Cruz, who from 2009 to 2017 chaired a committee advising STScI on Webb’s future science operations.

No, this is a horrible idea that will hurt the quality of work done. The data also become fully and freely available after a year. However, these scientists put in HUGE amounts of work to design and create their proposals and observations, and everything involved. They should get first pick at the data. This isn't just spending 5 minutes picking what you're gonna point the telescope it. Its months to years of work,

I disagree.  Science (as opposed to careers) works far better when data is revealed as soon as possible.   The current process often takes years for what could be done in weeks.  Scientist A makes an observation, B reads the paper a year later (and you can bet that if the proprietary period is a year, a year is what it will take) and has an interpretation/idea and publishes that another 6 months later, then that inspires C with a better theory, ands so on.  Conversely, as we saw during the pandemic, when everyone publishes everything they know as soon as they suspect it,  progress is extremely fast.

"Scooping" as a fear is over-rated, in my opinion.  It's actually good for science, getting the obvious results out faster, but folks (especially grad students) worry about the effect on careers.  However, in all cases I am personally aware of (I know this is anecdotal) what has happened is a senior researcher was sitting on a huge mound of data, intending to publish a giant paper once all detailed analysis is complete, and when nothing more can be wrung from the data.  But a younger and hungrier researcher saw there were individual points in this large data set that would make a good smaller paper by themselves, and published first.  The senior scientist's career was in no way ruined (they were already well known in their field) but they were annoyed, because even well established scientists love credit. But it was the senior scientist's idea to write a giant article in the first place, delaying the whole field in search of personal (or group) credit.  I suspect this scenario is particularly common with JWST data, since you're probably already a senior scientist if you can even get observing time on JWST.

[Dizzy_RHESSI]

The idea of scrapping proprietary time is misguided. It does seem to be conflating results and data. It would benefit big teams and I suspect it would have a negative effect on the science.  Look at the results of the ERS programs, GLASS and CEERS have generated a lot of papers but many of these were fired out as quickly as possible in order to be first. These collaborations did not take the time to (e.g.) properly set the flux calibration because they were under pressure to publish, and some results have changed substantially with newer calibrations. The methods being used were also not advanced, the tools were the same as from those from 10 years ago with HST and Spitzer, because it was fast. If every program was like this it would be chaos. Anyone trying to take the time to actually understand the data properly would be penalised.

It's also very reactionary to base it all on Cycle 1/Cycle 2, which this will have no effect on. In 2 years when the archive is dominated by public data this will be totally different. The ERS programs were successful in getting a bit of data for people to learn the ropes, but that doesn't mean the field should be turned upside down.

I think if you want to promote open science in a positive way they should focus on publications. I suspect the survey will get a very negative reaction, I haven't heard any astronomer defend this. It's also shortsighted to focus only on raw data, maybe you can reproduce someones analysis but it might take weeks or months. A better route would be to encourage teams to release high level data products (reduced images/spectra, catalogs) rather than just the archive publishing the bare minimum. A lot of papers analysed the three high redshift NIRSpec galaxies from the ERO data. There would have been far fewer papers if the commissioning team only released the raw data instead of their extracted spectra.

As it says in the article the period should be reduced to six months. A year is an excess amount of time to hold the data back for a publicly funded instrument.

And so you have your Cycle 1 program approved. You submitted the proposal 2 years ago. Half your data is taken in August, the other half is scheduled for August 2023. The first observations will be public 6 months before you have the complete data. So do you publish a half-baked analysis of half the data, just to avoid being scooped, or do you wait? Observers have no control over the schedule. Also bear in mind that some of the people currently getting data have been waiting for 3 decades.

It can be. In other cases, the suggestion is obvious and anyone in the field would have proposed it when asked "what should we do with this big IR bucket?" I think the mistake is having a one-size-fits-all embargo...

Submitting a proposal for JWST is not just writing a justification like other facilities, the full observing program down to each dither and exposure is required. It is not something that can be done in 5 minutes, even for a small program it is a week of work if you are experienced with APT and already have a plan. I doubt one could even configure the exposure time calculator alone in 5 minutes. It's also not something that people should be encouraged to take lightly, small differences in observing strategy can lead to large differences in overheads and the total time needed. Carefully optmising the strategy is how you maximise the scientific return and efficiency. And minimum effort proposals are pretty likely to be rejected. The current system is not "one-size fits all". The proposers select whether they want 12, 6, 3 or zero months proprietary time. 12 months is simply the default for small and medium, zero for large and pure parallel programs.

[eeergo]

There's a certain set of particularities that regards complex, bleeding-edge scientific observations made through instrumentation that took years, if not decades, to build *and characterize*, or whose calibration is still so rough that not even the teams in charge have a clear idea of how much further work is needed to have stable interpretations of the data - which is how most of modern science is done nowadays.

These peculiarities are usually exacerbated by the measurements being irrealizable without reliance on very significant teamwork, and the need for expertise only possessed by the owner teams.

To de-abstract these statements with something more tangible from first-hand experience:

I've worked in several large experimental collaborations focused on physics research. It's pretty easy to find numerous examples online, but to speak of what I know in detail about: I've seen preliminary, rough, qualitative results previewed at a presentation in a conference, a few weeks prior to publication of a peer-reviewed paper that was delayed for purely editorial reasons, ripped from any context, taken at face value, and single/few-author papers published before the "real" one by people outside the experimental collaboration claiming observation of, let's say, imaginative stuff. In case of questions about the rigorousness of the data, please contact the collaboration, fault is theirs, of course.

Evidently, their "results" were BS, in part because of the shoddy, rushed work, but mainly because of that missing context (calibrations, comparison with simulations relying on obscure empirical detector parameters...). Yet they managed to create controversy, rightfully piss people off, erode morale among those busting their asses to make the experiment work, or who spent years fine-tuning their analysis to minimize errors, and generally rob luster from the otherwise important, peer-reviewed observation. Moreover, a small part of these "studies" were misguided attempts at making science - yet most were dishonest attempts at gaining notoriety and, hopefully, get $$$ from funding agencies while the ruse held.

This is clearly an extreme example that borders on piracy - as such its end result is pretty straightforward. However, it can be easier to detect than having everyone and their grandma jump on raw data the moment it's generated, and rely they come up with rigorous results. Much easier to spew grandiose blurts than to refute them.

For JWST, we've already seen results of that "gold rush", even with the current data release policy: a mad dash to fame trying to find "the farthest galaxy" (whose title would probably have an expiry date close to that of a lettuce, even if the measurement was spot on, as soon as new images were available), resulting in embarrassing instantaneous retractions after making sensationalistic headlines. When this happens for just a few high-profile results, critiques can be fast and actually more constructive than with a more restricted data access policy. When this happens across the board though, the S/N ratio plummets, and confusion ensues as it's materially impossible to follow through every "inspired" claim.

[Welsh Dragon]

^ I couldn't agree more. All that immediate data availability is going to do is create a rush of sub-standard sloppy work.

Note that the argument "But it's tax-payers money!" is complete hogwash. People seem to think money just gets handed out to anyone who asks nicely, and so people are entitled to see everything immediately. No, money gets handed out after a lot of hard work, vicious competition and often several failed bids (yes there are issues and there is bias, but that's not the point). That work needs to be rewarded or there is no point to it.

In no other field that I am aware of is the raw data immediately available (if at all), regardless who funds the work. The resulting publications will be open access and the curated and validated data might be deposited somewhere. Astronomy is already more than generous in releasing all the data after an embargo period.

[eeergo]

IMHO there's a big difference between rapid/automated release of multipurpose, immediately interpretable data (the most obvious case are pretty pictures, but not only), or general documentation within proprietary or security limits (which should be greatly rewritten to cut excessive withholding, IMHO), which has shown to be of great value to both improve scientific output, public engagement and educational purposes - and indiscriminate immediate release of every bit of raw data from any instrument.

They are not the same, nor should they be, nor should anyone aim to bring them closer together, just as reasonable copyright laws for novel artistic works should not be treated on equal footing with charging a grandma for singing to her nephew a 200-year-old birthday song reinterpreted by Disney in one of their movies.

[Star One]

The idea of scrapping proprietary time is misguided. It does seem to be conflating results and data. It would benefit big teams and I suspect it would have a negative effect on the science.  Look at the results of the ERS programs, GLASS and CEERS have generated a lot of papers but many of these were fired out as quickly as possible in order to be first. These collaborations did not take the time to (e.g.) properly set the flux calibration because they were under pressure to publish, and some results have changed substantially with newer calibrations. The methods being used were also not advanced, the tools were the same as from those from 10 years ago with HST and Spitzer, because it was fast. If every program was like this it would be chaos. Anyone trying to take the time to actually understand the data properly would be penalised.

It's also very reactionary to base it all on Cycle 1/Cycle 2, which this will have no effect on. In 2 years when the archive is dominated by public data this will be totally different. The ERS programs were successful in getting a bit of data for people to learn the ropes, but that doesn't mean the field should be turned upside down.

I think if you want to promote open science in a positive way they should focus on publications. I suspect the survey will get a very negative reaction, I haven't heard any astronomer defend this. It's also shortsighted to focus only on raw data, maybe you can reproduce someones analysis but it might take weeks or months. A better route would be to encourage teams to release high level data products (reduced images/spectra, catalogs) rather than just the archive publishing the bare minimum. A lot of papers analysed the three high redshift NIRSpec galaxies from the ERO data. There would have been far fewer papers if the commissioning team only released the raw data instead of their extracted spectra.

As it says in the article the period should be reduced to six months. A year is an excess amount of time to hold the data back for a publicly funded instrument.

And so you have your Cycle 1 program approved. You submitted the proposal 2 years ago. Half your data is taken in August, the other half is scheduled for August 2023. The first observations will be public 6 months before you have the complete data. So do you publish a half-baked analysis of half the data, just to avoid being scooped, or do you wait? Observers have no control over the schedule. Also bear in mind that some of the people currently getting data have been waiting for 3 decades.

It can be. In other cases, the suggestion is obvious and anyone in the field would have proposed it when asked "what should we do with this big IR bucket?" I think the mistake is having a one-size-fits-all embargo...

Submitting a proposal for JWST is not just writing a justification like other facilities, the full observing program down to each dither and exposure is required. It is not something that can be done in 5 minutes, even for a small program it is a week of work if you are experienced with APT and already have a plan. I doubt one could even configure the exposure time calculator alone in 5 minutes. It's also not something that people should be encouraged to take lightly, small differences in observing strategy can lead to large differences in overheads and the total time needed. Carefully optmising the strategy is how you maximise the scientific return and efficiency. And minimum effort proposals are pretty likely to be rejected. The current system is not "one-size fits all". The proposers select whether they want 12, 6, 3 or zero months proprietary time. 12 months is simply the default for small and medium, zero for large and pure parallel programs.

All I will say in response is other NASA space telescopes have I believe a six month proprietary data period why should JWST be different.

[Greg Hullender]

It seems like just yesterday that the big debate was over whether scientists receiving public funds ought to ever have to release their data. I actually lobbied my congressman to support language in a bill that would force the eventual release of data--back when that was controversial.

I think a year is a reasonable amount of time. Yes, for something like COVID it probably makes sense to require immediate release, owing to the urgency of the issue. But it's hard to see where that urgency is coming from when it comes to pictures of galaxies. It took billions of years for that light to get here. Another year or two won't make much difference. :-)

[ttle2]

As said before, large GO programs already have all their data immediately available (it's possible to apply for an exclusive access period, but at least in Cycle 1 none of the large programs do that). DDR and other special programs obviously also have no exclusive access. So it's only the small and medium GO (and GTO, but that's a small fraction of time and it will be done by Cycle 3 anyway I think) that have a default of 12 months, which the proposal may choose to shorten or waive altogether (and maybe ~20% of proposers in Cycle 1 did so). While it's true that currently HST and Chandra have six month proprietary periods, that's a recent change and comparing JWST, which hasn't even issued Cycle 2 call for proposals, to HST and Chandra in Cycle 20+ is really apples and oranges. I do expect that also JWST small and medium program default proprietary period will be shortened in a few cycles time.

Secondly, I agree that the benefits of immediately available data is not that great. Unlike COVID research, astronomical observations are mostly not that time critical and for those that are, there's DDR. Unfortunately, a lot of bad flag-planting papers get written just to get priority on some discovery (and the mad COVID research rush did produce a lot of junk, too). The early "highest z" race was a good (bad') example of that and another example is the infamous BICEP2 data scraping from slides.

[matthewkantar]

If astronomers want exclusive use of the data, maybe they want to pay the prorated cost of their observation time? Otherwise the data should be public. Get a grip. Demanding an embargo on a freebie is gross.

[leovinus]

If astronomers want exclusive use of the data, maybe they want to pay the prorated cost of their observation time? Otherwise the data should be public. Get a grip. Demanding an embargo on a freebie is gross.

Sorry but you are wrong. Doing proper analysis and science takes takes time. Writing an astronomy grant, forming a team, defending all plus the analysis takes time. Give the scientists a break please. This is not yet-another-hear-me-shout-from-the-rooftops-I-am-entitled social media noise competition.

If it helps, we had similar discussions about “free” software decades ago. All software must be free! Share all! Blimey, when you run a business you need to make money. When you want to do a Richard Stallman, well you know where to go Or, medical research anyone? Paid by the public? Is all data free? Let’s also dissolve all patents as everything must be free? Thanks for giving me a smile.

In today’s society, tax payer funded research will  lead to shared results, papers, patents and data when possible but the historical precedents are clear that demanding free everything is silly.

[Lee Jay]

If astronomers want exclusive use of the data, maybe they want to pay the prorated cost of their observation time? Otherwise the data should be public. Get a grip. Demanding an embargo on a freebie is gross.

It's not a freebie.

Writing a proposal is usually unfunded, risky, and extremely time consuming. Getting observing time is just the start. You also need to find funding for the team, the analysis and the publication of results.  That's many hundred hours of unfunded proposal writing followed (if you are selected) by many thousands of hours of work.

That freebie probably cost in the seven figures range, excluding the scope and oberservations and a good bitnof that was probably unfunded.  I remember spending all of Thanksgiving in a family member's basement writing a proposal that ended up costing $75k to write not including all my nights, weekends and holidays, which added up to around 400 hours.

[Welsh Dragon]

If astronomers want exclusive use of the data, maybe they want to pay the prorated cost of their observation time? Otherwise the data should be public. Get a grip. Demanding an embargo on a freebie is gross.

This is exactly what I was talking about. Calling research funding a freebie shows a profound misunderstanding of how science funding works, and is deeply insulting to the scientists involved.

[Dizzy_RHESSI]

 

All I will say in response is other NASA space telescopes have I believe a six month proprietary data period why should JWST be different.

Yes it's quite different. JWST is a brand new telescope with dozens of modes, all of which are very poorly understood. The analysis pipeline is changing every week as new bugs are found, the calibrations are also changing day by day. There are instrumental artifacts in the data which the reduction software doesn't even deal with yet (e.g. 1/f noise, persistence). There are effects which were not accounted for in the instrument simulations (reflections, stray light, snowballs). It is quite different to HST were the instruments haven't changed for more than 12 years, with a stable pipeline. A Cycle 1 JWST program is much more work because you have to solve these problems yourself, and often there is no benchmark for you to compare to. Furthermore the pressure is much lower, partly because it has years worth of archive and partly because the low hanging fruit is gone. There is a huge experience gap, decades compared to months. 12 months was the default on HST for decades, it was fine. I think 12 months should have remained an option for HST and Chandra.

I'd also make a general point that JWST wasn't built to make data for it to just sit in an archive. JWST produces data, not scientific results. Astronomers do that. Take a real Cycle 1 example, the public program targeting TRAPPIST-1 with transit spectroscopy. Several epochs have been sitting in the archive for months available to anyone. Does the public benefit from this? No. Does this raw data accomplish any of JWST's science goals? No. But still people are excited to see these results, once the team actually publishes them. Reducing the data and extracting the results can be just as complicated as designing instruments. It's a lot of work, but without that the whole endeavor is pointless. The public engagement on JWST has been very nice with people processing images into pretty pictures, but most of the data isn't photogenic and ultimately it's not a 10 billion dollar wallpaper machine. The relationship between observatories and astronomers is symbiotic. Astronomers want to use the best instruments to advance the field. Telescopes want to attract the best talent to get the most out of the facility. It is not one way as some would suggest.

[LouScheffer]

All I will say in response is other NASA space telescopes have I believe a six month proprietary data period why should JWST be different.

Yes it's quite different. JWST is a brand new telescope with dozens of modes, all of which are very poorly understood. The analysis pipeline is changing every week as new bugs are found, the calibrations are also changing day by day. There are instrumental artifacts in the data which the reduction software doesn't even deal with yet (e.g. 1/f noise, persistence). There are effects which were not accounted for in the instrument simulations (reflections, stray light, snowballs). It is quite different to HST were the instruments haven't changed for more than 12 years, with a stable pipeline. A Cycle 1 JWST program is much more work because you have to solve these problems yourself, and often there is no benchmark for you to compare to.

This is an reasonable point.  However, some fraction of this time will not be astronomers themselves developing new calibration techniques, but simply waiting for the pipeline folks (who do not nearly get enough credit, in my opinion) to adjust their techniques and methods to the new machine.   In the meantime, in my experience, the astronomers are developing what results they can from the less-accurately calibrated data.  Others could do this as well.

I'd also make a general point that JWST wasn't built to make data for it to just sit in an archive. JWST produces data, not scientific results. Astronomers do that. Take a real Cycle 1 example, the public program targeting TRAPPIST-1 with transit spectroscopy. Several epochs have been sitting in the archive for months available to anyone. Does the public benefit from this? No. Does this raw data accomplish any of JWST's science goals? No. But still people are excited to see these results, once the team actually publishes them.

TRAPPIST-1 is actually an argument AGAINST proprietary periods.  The raw data was publicly released right away, yet no one got scooped, no rushed and shoddy papers came out (to my knowledge), and people will still be excited to see the correctly processed and calibrated results when they come out.   I think this will be the rule and not the exception, as scientists are quite capable of recognizing and being appropriately skeptical of preliminary work that took shortcuts.  (The "highest Z" race was an exception, I think, as the target was the media and not the scientific community, so primacy was more important than accuracy.  JWST will soon be past this point, if it's not already.)

Several epochs have been sitting in the archive for months available to anyone. Does the public benefit from this? No.

I don't agree with this.  The public benefits from this even if no member of the public ever looks at it.  It boosts confidence, not because a random member of the public has any ability to analyze it, but because it shows that other experts have had a chance to analyze it and declined.  As an extreme example, suppose you are giving a public talk on TRAPPIST-1 and some guy (it's almost always a guy) asserts the spectral data demonstrates the existence of aliens.  If you simply and correctly state "Our analysis shows no evidence of aliens", there will still be doubts, in him and others. (Are you doing the right analysis?  Are you hiding something because it's non-orthodox?  Etc.)  But if you can add that "the data is publicly available" then the questioner knows that the minority of experts who are pro-alien (Avi Loeb, etc.) have had a chance to look at it.  They would not be shy pointing out the obvious if it existed in the data.

It's much like political candidates releasing their income tax returns.  Are you going to read one of these yourself?  Probably not.  But you know that people who *are* experts in accounting will take at a look, and let you know if anything interesting pops out.

[deadman1204]

All I will say in response is other NASA space telescopes have I believe a six month proprietary data period why should JWST be different.

Yes it's quite different. JWST is a brand new telescope with dozens of modes, all of which are very poorly understood. The analysis pipeline is changing every week as new bugs are found, the calibrations are also changing day by day. There are instrumental artifacts in the data which the reduction software doesn't even deal with yet (e.g. 1/f noise, persistence). There are effects which were not accounted for in the instrument simulations (reflections, stray light, snowballs). It is quite different to HST were the instruments haven't changed for more than 12 years, with a stable pipeline. A Cycle 1 JWST program is much more work because you have to solve these problems yourself, and often there is no benchmark for you to compare to.

This is an reasonable point.  However, some fraction of this time will not be astronomers themselves developing new calibration techniques, but simply waiting for the pipeline folks (who do not nearly get enough credit, in my opinion) to adjust their techniques and methods to the new machine.   In the meantime, in my experience, the astronomers are developing what results they can from the less-accurately calibrated data.  Others could do this as well.

I'd also make a general point that JWST wasn't built to make data for it to just sit in an archive. JWST produces data, not scientific results. Astronomers do that. Take a real Cycle 1 example, the public program targeting TRAPPIST-1 with transit spectroscopy. Several epochs have been sitting in the archive for months available to anyone. Does the public benefit from this? No. Does this raw data accomplish any of JWST's science goals? No. But still people are excited to see these results, once the team actually publishes them.

TRAPPIST-1 is actually an argument AGAINST proprietary periods.  The raw data was publicly released right away, yet no one got scooped, no rushed and shoddy papers came out (to my knowledge), and people will still be excited to see the correctly processed and calibrated results when they come out.   I think this will be the rule and not the exception, as scientists are quite capable of recognizing and being appropriately skeptical of preliminary work that took shortcuts.  (The "highest Z" race was an exception, I think, as the target was the media and not the scientific community, so primacy was more important than accuracy.  JWST will soon be past this point, if it's not already.)

Several epochs have been sitting in the archive for months available to anyone. Does the public benefit from this? No.

I don't agree with this.  The public benefits from this even if no member of the public ever looks at it.  It boosts confidence, not because a random member of the public has any ability to analyze it, but because it shows that other experts have had a chance to analyze it and declined.  As an extreme example, suppose you are giving a public talk on TRAPPIST-1 and some guy (it's almost always a guy) asserts the spectral data demonstrates the existence of aliens.  If you simply and correctly state "Our analysis shows no evidence of aliens", there will still be doubts, in him and others. (Are you doing the right analysis?  Are you hiding something because it's non-orthodox?  Etc.)  But if you can add that "the data is publicly available" then the questioner knows that the minority of experts who are pro-alien (Avi Loeb, etc.) have had a chance to look at it.  They would not be shy pointing out the obvious if it existed in the data.

It's much like political candidates releasing their income tax returns.  Are you going to read one of these yourself?  Probably not.  But you know that people who *are* experts in accounting will take at a look, and let you know if anything interesting pops out.

If everything was free, you wouldn't have anything. How would a company get the money to design a car and build it if everything was free?

If the team that puts hundreds and hundreds of hours (unpaid) into getting an observation has no rights to the data, no one is going to plan the complicated important observations. Which means that they DON'T HAPPEN.  JWST wouldn't produce amazing science cause no team is going to put thousands of hours into the work only to be scooped by some random person who gets much of the data wrong, but steals the credit anyways.

This argument would have some validity if the data was never made public. However it IS made public. The good that the data embargo does has been very clearly stated over and over again. There has been zero attempt to address all the harm removing the data embargo would do. Just "but free cause I didn't do anything for it and want it now". Actions have consequences, and such an action needs to justify itself.

[edzieba]

If everything was free, you wouldn't have anything. How would a company get the money to design a car and build it if everything was free?

JWST is not a car, and nobody is making the strawman argument of "everything everywhere should be free!".

If the team that puts hundreds and hundreds of hours (unpaid) into getting an observation has no rights to the data, no one is going to plan the complicated important observations. Which means that they DON'T HAPPEN.  JWST wouldn't produce amazing science cause no team is going to put thousands of hours into the work only to be scooped by some random person who gets much of the data wrong, but steals the credit anyways.

And yet for telescopes and other instruments without embargo periods, observations continue and instrument time continues to be oversubscribed. 'Data available immediately means nobody will take the effort to apply for observation time' does not seem to have ever been the case anywhere. I suspect you'd also be hard-pressed to find a scientist who, faced with the opportunity to be the first to make a discovery using an instrument with a brand new capability, would not choose to take up the challenge of doing so faster than others but instead actively choose to remain in ignorance due the risk of being second to know.

[JayWee]

(The "highest Z" race was an exception, I think, as the target was the media and not the scientific community, so primacy was more important than accuracy.  JWST will soon be past this point, if it's not already.)

Tbh, people will rememeber only "JWST spotted galaxy 13.8 billion years old". And then have no idea who the people who calculated it are. No point in primacy.

[TrevorMonty]

We've been waiting decades for overdue JWST, another year to access free data isn't big deal.

[deadman1204]

If everything was free, you wouldn't have anything. How would a company get the money to design a car and build it if everything was free?

JWST is not a car, and nobody is making the strawman argument of "everything everywhere should be free!".

If the team that puts hundreds and hundreds of hours (unpaid) into getting an observation has no rights to the data, no one is going to plan the complicated important observations. Which means that they DON'T HAPPEN.  JWST wouldn't produce amazing science cause no team is going to put thousands of hours into the work only to be scooped by some random person who gets much of the data wrong, but steals the credit anyways.

And yet for telescopes and other instruments without embargo periods, observations continue and instrument time continues to be oversubscribed. 'Data available immediately means nobody will take the effort to apply for observation time' does not seem to have ever been the case anywhere. I suspect you'd also be hard-pressed to find a scientist who, faced with the opportunity to be the first to make a discovery using an instrument with a brand new capability, would not choose to take up the challenge of doing so faster than others but instead actively choose to remain in ignorance due the risk of being second to know.

Please list the observatories that are as complicated to use and analyze the data from as JWST is without an embargo period on any data.

I'll help you out -  any observatory thats been running for several years WON'T be as complicated to use and analyze as JWST, because they will be well understood.

[ttle2]

Please list the observatories that are as complicated to use and analyze the data from as JWST is without an embargo period on any data.

I'll help you out -  any observatory thats been running for several years WON'T be as complicated to use and analyze as JWST, because they will be well understood.

Or any highly subscribed observatories with no default proprietary period. The ones I checked (HST, Chandra, XMM Newton, Integral, ALMA, VLA, VLT, Keck, Gemini) all still seemed to have a proprietary period of at least six months for normal GO programs.

TRAPPIST-1 is actually an argument AGAINST proprietary periods.  The raw data was publicly released right away, yet no one got scooped, no rushed and shoddy papers came out (to my knowledge), and people will still be excited to see the correctly processed and calibrated results when they come out.   I think this will be the rule and not the exception, as scientists are quite capable of recognizing and being appropriately skeptical of preliminary work that took shortcuts.  (The "highest Z" race was an exception, I think, as the target was the media and not the scientific community, so primacy was more important than accuracy.  JWST will soon be past this point, if it's not already.)

The "highest z" race was definitely not just for the public. The early papers were submitted (and at least one accepted) to top journals: Nature, ApJL, MNRAS...