Talk:List of Falcon 9 and Falcon Heavy launches/Archive 8

RfC on booster landings graph
Should the booster landings graph be simplified? More specifically, is the graph difficult to understand due to having too much detail? A closely related issue is that most editors agree that the captions "Ocean failure", "Parachutes failure", and "Ocean touchdown" should be improved, but we cannot come to a consensus as to how. Simplifying the graph would solve this.

The options for simplification are:

Tercer (talk) 15:14, 7 January 2020 (UTC)
 * Option 1: Do nothing.
 * Option 2: Merge the categories "Ocean failure", "Parachutes failure", and "Ocean touchdown" together into "Test (no recovery)".
 * Option 3: Merge the categories "Ocean failure", "Parachutes failure", "Ocean touchdown", and "No attempt" together into "No attempt".
 * Option 4: Same as option 2, and also merge both kinds of success into "Success" and both kinds of failure into "Failure".

The simplified graphs look like this:

Poll

 * Oppose changing the categories as presented. I could perhaps support modifying the captions for the parachutes and ocean options, but that hasn't been presented as an option here. Additionally, Option 1 is the only one that points out the distinction between successful ocean landings and failed ones (thanks for for reminding me of this). —  Insertcleverphrasehere (or here)  (click me!)    19:54, 7 January 2020 (UTC)
 * Oppose There is a massive difference between successful ocean landing and failed ocean landing. AndrewRG10 (Talk) - 20:43 7 January 2020 )UTC) —Preceding undated comment added 20:44, 7 January 2020 (UTC)
 * Oppose changes to the chart, but improving the captions would be a good idea. I'm not strongly opposed to changing the chart, but we have discussed this and there was only one editor favoring a change. Fcrary (talk) 22:21, 7 January 2020 (UTC)
 * Oppose. I disagree with the way the poll is introduced as it makes blatantly wrong claims. If I would consider this a legitimate poll then I would favor option 1, don't change it. Changing labels could be interesting if someone has specific proposals. --mfb (talk) 23:55, 7 January 2020 (UTC)
 * Oppose changing the categories as presented. Per discussion below, captions don't seem confusing nor do I think there are too many categories. crandles (talk) 00:01, 8 January 2020 (UTC)
 * Keep as is – Looks clear enough, and plenty of sources are available for readers curious to learn more. — JFG talk 01:21, 8 January 2020 (UTC)
 * Oppose the current proporal. I support the following proposal:




 * --Soumyabrata (talk • subpages) 09:05, 8 January 2020 (UTC)
 * Soumyabrata, I don't understand what your proposal is, looks like there was some problem with the code. Isn't it the same as Option 2?
 * I prefer Tercer's 'Test (No recovery)' caption per option 2 to just 'No recovery'. No significant view on colour change. Are there any other differences? But as others say whether test is a success or not seems an important detail if we are showing the tests. crandles (talk) 10:33, 8 January 2020 (UTC)
 * This is exactly the Tercer's option 2 chart. Aside color, the only difference from the option 2 is the CRS-16 is included in the "No recovery", rather than "Ground-pad failure". The reason for this is that the CRS-16 successfully touched down on the Atlantic Ocean. --Soumyabrata (talk • subpages) 17:39, 8 January 2020 (UTC)
 * I don't agree with putting it in that category, they were attempting a ground pad landing so it should be in 'Ground pad failure' category. The 'No recovery' is ambiguous it might mean successfully landed but not recovered (eg ArabSat-6A) in addition to, or instead of, ocean tests. Therefore, caption needs to be 'Test (no recovery)' and then CRS-16 clearly doesn't fit in that category. crandles (talk) 18:21, 8 January 2020 (UTC)
 * Thanks for illustrating how unclear the current categories are. The category "ocean touchdown" is not meant to include all boosters that performed a successful ocean touchdown, but only those which did it when that was the goal. The goal in the CRS-16 case was to land on the ground-pad, so it belongs unambiguously in the "ground-pad failure" category. I named the category "Test (no recovery)" to make it clear that only the landing tests belong there, not merely the boosters that were not recovered. Tercer (talk) 18:23, 8 January 2020 (UTC)
 * My current chart is now exactly the option 2 chart with some modifications. The CRS-16 is now included in the "Ground-pad failure". --Soumyabrata (talk • subpages) 13:46, 9 January 2020 (UTC)
 * , to clarify, this is essentially Support option 2 with a few cosmetic changes. —  Insertcleverphrasehere (or here) (click me!)    22:52, 12 January 2020 (UTC)


 * Oppose The current graph could be improved, but it's superior to any of the options prevented. – Jadebenn (talk &middot;&#32;contribs &middot;&#32;subpages) 05:39, 9 January 2020 (UTC)
 * It would be helpful if you could specify how you think the graph could be improved. Tercer (talk) 05:55, 9 January 2020 (UTC)


 * Oppose we lose the evolution of experiments. There is a massive difference between successful ocean landing and failed ocean landing. --Dwalin (talk) 18:06, 9 January 2020 (UTC)

Discussion
It is absolutely incorrect that "all editors agree that the captions "Ocean failure", "Parachutes failure", and "Ocean touchdown" are confusing or misleading" this is a gross mischaracterisation of the discussion section above. Several people said that they could see improving the captions for those categories, that's all. In any case the purpose of this RFC seems to be changing the categories, rather than fixing any issue with the captions, (which is confusing given that it was the thing that actually had some support in the previous discussion). The collection of categories in the current graph represents the information contained in the list itself and should be retained. The other options do not. Options 2 and 4 seem to be equating 'landing' with 'recovery' and confusing the two (remember that this graph is captioned "Booster landings"). Option 3 is right out, indicating that 'No Attempt' was made at landing, when Ocean landings were attempted (remember that especially in the first couple ocean landings, they were planning to see if they could recover the booster after it landed in the ocean and towing it back to shore). —  Insertcleverphrasehere (or here) (click me!)    19:54, 7 January 2020 (UTC)
 * I think "all editors agree that the captions "Ocean failure", "Parachutes failure", and "Ocean touchdown" are confusing or misleading" is a fair summary of the discussion above (anyone can read for themselves), but that's not important, I'd be happy to change it to "should be improved", if it is possible to change a RfC statement (I don't know).
 * Indeed the purpose of the RfC is to simplify the categories, not change the captions. I'm just mentioning that solving the caption problem is a nice side-effect of the simplification.
 * About "confusing" landing and recovery: you seem to be using a very strange definition of landing, that would include the parachuting down of the Shuttle SRBs into the ocean. This is what is meant by the category "Parachutes failure", which I doubt anybody would be able to guess. Note that this is not the definition used in the page Falcon 9 first-stage landing tests, for example. I think any reader would think of landing as bringing down the booster to rest safely in a solid surface (as evoked by "land"). Using "landing" to include making a soft ocean touchdown is confusing and misleading. Tercer (talk) 20:25, 7 January 2020 (UTC)
 * , Well, the ocean landing tests were retro-propulsive landings at the surface of the water. They are included in the Falcon 9 first-stage landing tests article, so they are 'landing tests'. They are referred to as 'landings at sea' in the SpaceX reusable launch system development program article. Most importantly, the list in this article lists them under the 'Booster landing' column. I can appreciate that there is room to argue that the ocean landings should not be referred to as 'landings' but your definition does not seem to be the one used amongst the various wikipedia articles we have on this subject. —  Insertcleverphrasehere (or here) (click me!)    21:09, 7 January 2020 (UTC)
 * Insertcleverphrasehere, actually the article Falcon 9 first-stage landing tests calls them "ocean touchdown attempts", as opposed to the "landing attempts". They do belong in the landing tests page, as they were tests that led to the development of landing. It is not true that they are referred to "landings at sea" in the SpaceX reusable launch system development program article. Those refer to drone-ship landings. The ocean touchdowns are referred there as "descent and simulated landing tests". Furthermore, I don't see how the fact that this article uses this definition (presumably written by you) is an argument. That's precisely what I'm trying to change! This article is the only one using this misleading terminology. Tercer (talk) 21:25, 7 January 2020 (UTC)
 * , "presumably written by you"... um, I don't think so. This has been in effect for a long time, since very early in this list's development, well before it became a featured list, though there have previously been discussions I was involved in about the structure of list that might have led to it. I'm not sure and don't care to dig through the archives. It appears that I was mistaken about my statement about the SpaceX reusable launch system development program article, apologies. Not that it really matters. The definitions that you are trying to impose on this article really aren't an improvement over current convention, and would require a change in how the list is structured as well (or else result in a disconnect between the information shared in the list and that showed in the charts), which I, and I suspect most others that edit this page, would oppose. The fact that all your changes remove info about whether the ocean landings were successful or not is actually the primary issue, and one that can't be brushed aside with nebulous arguments about definitions. —  Insertcleverphrasehere (or here) (click me!)    23:07, 7 January 2020 (UTC)
 * Indeed, we should correct the list as well, as this bizarre definition also causes confusion there. And indeed, the caption problem is a side issue, what I want to discuss is the simplification of the graph. A point that you might have missed is that my goal is to remove information about whether the ocean touchdowns were successful or not. I don't think this information is relevant (the purpose of the test was not to perform a touchdown per se, the purpose was to develop landing), on the contrary, I think it is excessive detail that is making the graph unreadable. Tercer (talk) 23:24, 7 January 2020 (UTC)
 * , You seem to be the only one who thinks that information is irrelevant, I guess I can simply agree to disagree with you at this point. —  Insertcleverphrasehere (or here) (click me!)    23:27, 7 January 2020 (UTC)
 * Indeed, we all know that. That's why I started the RfC, to invite outside editors to give their perspective. In this way we can empirically determine if normal readers find this information relevant, and if they find the graph easy to understand. Tercer (talk) 23:32, 7 January 2020 (UTC)
 * , how can you think it's a fair summary if people tell you that they do not agree with it? Just to avoid any doubt here: I do not think that they are confusing or misleading. I don't claim they are the best possible captions ever, but so far I haven't seen better suggestions. --mfb (talk) 00:01, 8 January 2020 (UTC)
 * mfb, I already said that I'm happy to rewrite that sentence if one can do that in a RfC statement. Would you be ok with "should be improved"? Out of curiosity, what do you think is wrong with them, if they are not the best possible captions ever? Tercer (talk) 00:15, 8 January 2020 (UTC)
 * , no, I don't think we all agree that they "should be improved" either. I don't think there is something wrong with them either. I say it is possible that there are better captions, but so far I haven't seen a better suggestion. --mfb (talk) 00:21, 8 January 2020 (UTC)
 * "Could be improved" then? Tercer (talk) 00:27, 8 January 2020 (UTC)
 * , technically yes, but that applies to everything everywhere, pointing it out gives a wrong impression. mfb (talk) 05:01, 9 January 2020 (UTC)
 * How can I make you happy then? Tercer (talk) 05:19, 9 January 2020 (UTC)
 * You could stop trying to speak for everyone in general when you have an extreme minority opinion. mfb (talk) 05:34, 9 January 2020 (UTC)
 * Do you have a concrete suggestion on how to rephrase that sentence? Tercer (talk) 05:54, 9 January 2020 (UTC)
 * I think I made a very concrete suggestion. Don't make that sentence at all. mfb (talk) 06:57, 9 January 2020 (UTC)
 * That's not acceptable. There's clearly an issue with the captions. You're the only one that disagrees. Tercer (talk) 07:37, 9 January 2020 (UTC)
 * , That's (a) wrong and (b) even if I would be the only one it would still make a statement about every participant wrong. Please stop making claims about what I mean, because you have been getting it wrong pretty consistently now. Better stop making claims about what others mean in general, because you don't have a good track record there either. mfb (talk) 03:15, 11 January 2020 (UTC)
 * I've changed it to "most editors agree that the captions "Ocean failure", "Parachutes failure", and "Ocean touchdown" should be improved". Happy now? Could you also please remove your assertion that the poll is not legitimate and makes blatantly wrong claims? That's false and hurtful. Tercer (talk) 07:43, 11 January 2020 (UTC)

Could we possible stick to the point? This section of the talk page is about the chart and not the captions. Tercer feels that it is too complicated, contains too much information for the general reader and has made some suggestions to change that by simplifying the chart. I think it is very clear from the poll and the discussion that no one else agrees. Now, we have a second and separate issue of the captions. I think it would help if we stopped mixing together the discussions of those two separate issues. Let's close off the discussion of simplifying the chart with a "No" and open a new section on improving the captions. And, no, that doesn't mean I'm saying the captions are bad, just that they could be better. Fcrary (talk) 03:44, 11 January 2020 (UTC)
 * , Just use strikeout tags and make whatever changes you like. "could be improved" is probably more accurate. —  Insertcleverphrasehere (or here) (click me!)    00:22, 8 January 2020 (UTC)
 * Seeing the term parachute failure, I would expect most people to think an attempt was made to use parachutes and it failed. This doesn't seem confusing at all to me. 'Ocean touchdown' and 'ocean failure' perhaps have a little more scope for confusion but seem reasonably clear to me. I can see an argument for saying that in a year or two this detail may look slightly odd/out of place/more appropriate to an article on development of landing than this list. However for the moment, propulsive landing is a major, newish feature for this particular rocket and I don't see much problem with the number of categories or their descriptions. crandles (talk) 23:58, 7 January 2020 (UTC)
 * crandles, I think any normal reader would look at "parachute failure" and think that it was a landing attempt (it wasn't) that failed, perhaps because the parachutes didn't open (which is not the case). Similarly, any normal reader would look at "ocean failure" and think that it was a landing attempt (it wasn't) that failed, perhaps by hitting the ocean instead of the landing pad (which is not the case). "Ocean touchdown" sounds like a landing attempt (it wasn't) that ended in an ocean touchdown (which is true), but it is not clear whether the goal was an ocean touchdown or an actual landing. Tercer (talk) 00:22, 8 January 2020 (UTC)
 * , Fine... there is some ambiguity, but we can't avoid that. The solution isn't throwing the baby out with the bathwater and just saying that 'No Attempt" was made, or that there is no difference between a successful test landing and a failed one. —  Insertcleverphrasehere (or here) (click me!)    00:25, 8 January 2020 (UTC)
 * Of course we can avoid that, by not trying to present so much detail. It's a two-to-three word caption in a graph called "booster landings". Of course we can't convey complex information in this format. Tercer (talk) 00:31, 8 January 2020 (UTC)
 * , Aaaaand, that involves throwing the baby out with the bathwater. Let me make myself agonizingly clear so that perhaps you understand: to "avoid that" ambiguity we also lose something else; being able to show success/failure in Ocean Landings. No one else seems to consider this to be a net positive change. —  Insertcleverphrasehere (or here) (click me!)    02:10, 8 January 2020 (UTC)
 * How can't you understand that you are not managing to show that?! Nobody can look at a graph titled "booster landings", see a caption written "ocean failure", understand that it was an attempt at an ocean touchdown that failed, see a caption written "ocean touchdown", and understand that it was an attempt at an ocean touchdown that succeed. If the see a caption written "test" or something, at least they'll know that there was something different about that booster landing, and they can read the description in the list to understand what happened. Tercer (talk) 02:36, 8 January 2020 (UTC)
 * , It seems pretty clear to me. The booster attempted to land in the ocean. It failed, something went wrong. The booster made an ocean touchdown landing. Cool. Do you think our readers are too stupid to figure that out? Additionally, if they are confused they can look down at the list for more info and go, "oh... I understand that caption now, I'll head back to that graph and now I can have a look at how those landings/landing tests progressed over time in a nice graphical layout".
 * Perhaps there is room to improve the captions (I've made some suggestions, but we honestly haven't delved too deep into that), but throwing out the characterization of how the ocean landings went is not a good solution in my opinion. I get that you don't respect my opinion on this matter. That's fine. Please respect that mine is different and that consensus does not mean unanimity. —  Insertcleverphrasehere (or here) (click me!)    02:44, 8 January 2020 (UTC)
 * Of course you find it clear, you know this article inside out. The problem is that usually the reader is somebody that doesn't know the article inside out.
 * Look, when you put a figure in an academic paper, your goal is to make the figure self-contained, as people want to look at it and understand what's going on without needing to read the whole paper. That's often impossible, though, due to the complexity of the subject matter. Then the goal becomes to make it clear what in the figure is complex or unfamiliar, and point where in the paper it is explained. What you really want to avoid is to have some concept in the figure that the reader will think they understand, but they don't actually, because then the reader will not look for explanation and stay with the wrong concept. The graph we're talking about is squarely in the third category. Tercer (talk) 03:01, 8 January 2020 (UTC)
 * , Haha, I dug through the archives of this page. Turns out that I started out with the same view as you've got and was convinced otherwise. Have a look here: Talk:List_of_Falcon_9_and_Falcon_Heavy_launches/Archive_6 I think we can both get a good laugh out of that one. :D —  Insertcleverphrasehere (or here) (click me!)    04:43, 8 January 2020 (UTC)
 * Indeed, that's hilarious. If only you could be talked back into reason. What changed your mind, though? I looks like you simply deferred to JFG's opinion. Tercer (talk) 09:11, 8 January 2020 (UTC)
 * Didn't you complain about childish replies and personal attacks earlier? Implying everyone apart from you is unreasonable is not a good idea. mfb (talk) 05:06, 9 January 2020 (UTC)
 * , I respect 's opinion. He is an even more active editor of this page than I am, and his comments were good ones. I considered them, we ended up with a nice compromise that both of use could accept (his later proposal). Not every discussion has a winner and a loser. —  Insertcleverphrasehere (or here) (click me!)    18:11, 9 January 2020 (UTC)
 * I'm asking because I didn't see any argument about that. At that time you clearly thought that the graph was too complicated, but now you apparently don't anymore. didn't argue this point, he merely argued that the detail about the tests is important. This might as well be true, but it doesn't change the fact that the graph is too complicated. If both things are true, then the solution is too split the graph into two, one about the landing attempts, and another about the development of landing. The current graph mixes these two kinds of information, and the result is confusion. Tercer (talk) 19:57, 9 January 2020 (UTC)
 * , It may be a complicated graph; I don't dispute this. But if the detail is important, then it isn't "too complicated". —  Insertcleverphrasehere (or here) (click me!)    21:00, 9 January 2020 (UTC)
 * You're missing the point. Whether it is too complicated or not does not depend on the importance of the detail. If you put too much important information in any graph it will become too complicated. Whether it is too complicated or not depends only on whether the readers understand it. The (admittedly limited) empirical evidence that we have indicates that they do not. Tercer (talk) 07:33, 10 January 2020 (UTC)
 * Let's not underestimate our readers' capacity of understanding. The graph as displayed was a result of many discussions, in which clarity and complexity were debated and editors settled on a comprehensive presentation. The landing attempts are nicely categorized in a progression from parachute tests, ocean touchdown tests, to actual landing tests, further subdivided into ground-pad landings and drone-ship landings. Even the legends are neatly aligned with failures and successes of each landing mode presented opposite one another. If your remaining concern is the clarity of the legends, I would suggest adding a few explanatory footnotes, which can in turn point to the relevant sections of our dedicated article: Falcon 9 first-stage landing tests.
 * A side point about graph complexity in general: I would recommend reading Edward Tufte's The Visual Display of Quantitative Information and its sequels, in which Tufte makes the case that the best graphs do convey a lot of information thanks to numerous techniques such as axis choice, line thickness, subtle colors and well-crafted legends. — JFG talk 08:18, 10 January 2020 (UTC)
 * This is not about underestimating the reader. I have plenty of experience with experienced researchers (by all standards brilliant people) misunderstanding points in my papers that I thought were crystal clear. The answer to that is not to insult the reader and insist they should have understood it, but to rewrite the offending passage to clear up the confusion. Whether the readers understand this graph or not is an empirical question, that should be settled by empirical evidence, not by arguing amongst ourselves. And again, the (admittedly limited) evidence we have is that they do not. Tercer (talk) 09:05, 10 January 2020 (UTC)

Is there a case for changing the title from 'Booster landings' to something more like 'Booster landings and test progression'? Does that help make it clearer that some categories are about tests rather than actual landing attempts which seems to be 's issue? crandles (talk) 10:00, 8 January 2020 (UTC)
 * I don't think that would help, because you still couldn't guess from the captions what was a test and what was a landing attempt. To solve this problem I would simply leave the title as "Booster landings", and rename the test categories to "Test (parachute uncontrolled)", "Test (ocean uncontrolled)", and "Test (ocean controlled)", or something like that. I think that would make it clear that they were not landing attempts (although calling the parachute thingy a "landing test" is quite a reach). Tercer (talk) 19:10, 8 January 2020 (UTC)
 * I am thinking I would prefer 'Parachute test failure', 'Ocean test failure' and 'Ocean test controlled' as being shorter and without brackets and restricts use of success to actual landings but am willing to hear reasons why it should be different. crandles (talk) 22:33, 8 January 2020 (UTC)
 * I would avoid the word failure for the same reason you're avoiding the word success: preventing confusion with success or failure of an actual landing attempt. Tercer (talk) 22:42, 8 January 2020 (UTC)
 * I think there is a different problem with using uncontrolled in place of failure: An 'Ocean test uncontrolled' could be testing something other than control such as landing leg deployment and be considered a successful test. If we are showing tests then I want to know if they are successful or not. If it says 'Ocean test failure' then it is not that difficult to understand that you don't actually land on the ocean and it is a 'test failure', so obviously didn't work. Furthermore, the two ocean test categories are failure and controlled so it is clear which one is the more successful test. Whether you parse it as 'Ocean, test failure' or as 'Ocean test, failure' it is pretty clear it is a test not a landing that has failed. So I think the possibility of an uncontrolled test being successful is the bigger issue. Therefore, I am still preferring 'test failure' to using 'uncontrolled'. (It is also shorter and I am not sure if there is room for 'Ocean test uncontrolled') crandles (talk) 23:03, 8 January 2020 (UTC)
 * Why don't you want to use "Ocean test success" then? Anyway, I don't think it is at all clear. One might as well think that "Ocean test" meant a test landing at a drone-ship, as opposed to a regular landing.
 * What I find more problematic is that using the categories "success" and "failure" implies that the purpose of the test was the touchdown, which is not true. The first landing test, for example, was a great success, as they managed to reignite Merlin and do a controlled reentry. That was the purpose of the test, SpaceX wasn't trying to do a touchdown. Putting that as a "Ocean test failure" is misleading. For the landing attempts, on the other hand, it is clear cut: they tried to land, and they succeeded or failed. Tercer (talk) 23:35, 8 January 2020 (UTC)
 * Sounds like you are making a good case for that one, CASSIOPE, to be classed as 'Reentry test good'. I thought you wanted fewer categories not more. I wouldn't be too concerned about failure/uncontrolled especially if 'Ocean test uncontrolled' is more generally preferred to 'Ocean test failure'. So should we do a poll between different sets of captions? If so are there other people preferences besides: As is, my set with 'Ocean test failure', with 'Ocean test uncontrolled', option for 'Reentry test good'? crandles (talk) 14:07, 9 January 2020 (UTC)
 * Oh come on, now you're just making fun of me. Of course I don't want yet another category, I'm just pointing out that success/failure is not appropriate for the tests. I'm fine with a poll (without "reentry test good"). You didn't answer my question, though, why don't you want to use "Ocean test success"? Tercer (talk) 15:29, 9 January 2020 (UTC)
 * I said "I wouldn't be too concerned about failure/uncontrolled", it should be what is generally preferred. Together with previous answers as to what I preferred, how is this not answering your question? Are you any happier if I answer about success as I was trying to accommodate your wishes by keeping successes to actual landings. crandles (talk) 18:45, 9 January 2020 (UTC)
 * I just wanted to understand your reasoning, because the problem I see with success is the same problem I see with failure. But if you're just accommodating my wishes, fine, that's an answer. Tercer (talk) 19:45, 9 January 2020 (UTC)

Forget my previous proposal, I have changed my mind. I am currently proposing something like this (excluding the chart):

The "parachute failure" is merged with the "ocean failure", thus removing the excessive detail a bit. --Soumyabrata (talk • subpages) 17:31, 11 January 2020 (UTC)
 * If there is consensus to list fewer categories in this chart, then this particular grouping would make good sense. But I still favor the current setup. — JFG talk 17:49, 11 January 2020 (UTC)
 * , I think the parachute entry tests are wildly different from the failed retropropulsion ocean test landings. —  Insertcleverphrasehere (or here) (click me!)    17:55, 11 January 2020 (UTC)
 * There is no distinction between the retropropulsion splashdown and parachute splashdown in my proposed chart. According to Tim Dodd, "[splashdown] was probably the next step after successfully deploying the chutes [of Falcon 9]." --Soumyabrata (talk • subpages) 06:56, 12 January 2020 (UTC)
 * , Perhaps, obviously both types had some similarities, but our current setup helps illustrate the progression of different methods well, and I think we would lose something if we lumped those together. —  Insertcleverphrasehere (or here) (click me!)    07:22, 12 January 2020 (UTC)

Suggest SNOW close
Looks that there is no support for any of your suggested changes at this point. Would you consider a voluntary close of this RfC per WP:SNOW? The debate about legends can continue without touching the graphs themselves. — JFG talk 01:24, 8 January 2020 (UTC)
 * No. The explicit purpose of this RfC was to get a perspective from editors who are not SpaceX fans, as we're speculating about what they think. Apart from you (it turns out that JFG is the one who left the graph like this in the first place), all comments up to now were from the editors that were already arguing about the issue, bringing absolutely zero new information. Tercer (talk) 01:28, 8 January 2020 (UTC)
 * , did you advertise it anywhere else? I put a post on the WikiProject Spaceflight talk page and at Talk:Falcon_9, not sure where else we can go to find people who are keen to offer an opinion for an RfC about a technical detail of a single chart in an article with fringe interest. What did you expect? This isn't exactly a situation where the question sets some sort of precedent and demands widespread comment, it's just a case of one editor coming to a page with an established editor base and trying to change something because they WP:JUSTDONTLIKEIT, and hoping to find others that also don't like it. —  Insertcleverphrasehere (or here) (click me!)    01:57, 8 January 2020 (UTC)
 * , I've made a post at Village_pump_(miscellaneous). That's the best I can do. —  Insertcleverphrasehere (or here) (click me!)    02:15, 8 January 2020 (UTC)
 * I haven't advertised it. It's not a technical point, it's a matter of editing and design. My hope is that somebody who understands something about publishing will show up and tell you that it's atrocious design that would never see print in an academic journal, and you would listen. I am myself a professional physicist who has published dozens of papers, and I can tell you that this is a terrible way to present technical information, but it clearly doesn't matter what I say. Tercer (talk) 02:46, 8 January 2020 (UTC)
 * What you say does matter; all editors, new and old, experts and newbies, are welcome to contribute and debate. I hope you won't feel rebuked because most commenters here happen to disagree with your suggestions. Let's see what happens when more editors chime in. — JFG talk 10:07, 8 January 2020 (UTC)
 * I'm a physicist as well, I have reviewed and contributed to enough papers to lose track and I'm in the author list of hundreds of them (but that doesn't mean much in particle physics). This graph should be easy to read for physicists and I don't see why this would have a problem in a journal. You keep inventing new exotic interpretations of it that no one else ever seemed to have. mfb (talk) 07:04, 9 January 2020 (UTC)
 * Empirical data shows otherwise: the only user who is not an author of this article that commented about it misunderstood the graph. We don't know about Jadebenn, as they didn't try to interpret the graph or said whether they find it easy to understand. Tercer (talk) 07:42, 9 January 2020 (UTC)
 * that commented about it - everyone else was happy with the graph, or at least happy enough to not comment. For years. We know that Jadebenn prefers the current version over all alternatives you suggested. --mfb (talk) 07:52, 9 January 2020 (UTC)
 * You shouldn't assume that the data we don't have supports your position, that's rather unscientific. It might as well be that the reader wasn't happy about but wasn't bothered enough to complain, or, as was actually the case here, the reader thought it had understood it but actually didn't. Furthermore, the captions haven't been there for years, they were changed in 2018. Previously they were "Ocean controlled" and "Ocean uncontrolled". Tercer (talk) 08:20, 9 January 2020 (UTC)
 * Let's not get caught up in Warnock's dilemma, shall we? — JFG talk 08:23, 10 January 2020 (UTC)
 * If I may chime in, this was brought to my attention by the notice on WikiProject Spaceflight. My primary objection is the reclassification of recovery failures as either "test" or "no attempt" that occurs in all proposed versions of the graph. I feel that is an unnecessary muddying of the waters, and thus find the current graph superior. – Jadebenn (talk &middot;&#32;contribs &middot;&#32;subpages) 10:05, 9 January 2020 (UTC)
 * They were not recovery attempts, so I don't think it is accurate to call them recovery failures. Tercer (talk) 10:29, 9 January 2020 (UTC)
 * , That's also not entirely true. When they first attempted ocean landings they did attempt to see if they could be recovered by being towed back to shore, but the boosters back then tended to explode. (they actually did this a couple times later on, but mostly by accident, or because the block 5 was less prone to explosion). —  Insertcleverphrasehere (or here) (click me!)    10:35, 9 January 2020 (UTC)
 * They did try to recover the booster on flights 6,9, and 10, but not on flights 13, 15, 46, and 48. But that's a completely different issue, as the graph is about whether the touchdown was successful, not about whether the recovery was successful, and in any case I don't think that's what Jadebenn had in mind. Tercer (talk) 11:02, 9 January 2020 (UTC)
 * , My point is that trying to turn this into a 'recovery' graph has its own pitfalls. Which is why I abandoned this idea a long time ago. —  Insertcleverphrasehere (or here) (click me!)    18:02, 9 January 2020 (UTC)
 * I'm glad we agree on something, then. Tercer (talk) 18:40, 9 January 2020 (UTC)
 * I think it is important to distinguish between a failed attempt to land and not making an attempt. Currently, and for the foreseeable future, there are three possibilities. SpaceX attempted to land and did so, they attempted a landing and failed, and they did not attempt to do so because the customer's requirement precluded a landing attempt. To me, not trying because that's what the customer wanted is very different from trying and failing. Also, could we stop talking about recoveries? We had a lengthy discussion of this, over the booster that landed on a drone ship but was lost on the way back to the Cape due to sea conditions. We decided the landing was successful although the recovery was not. And that the chart was about landings not recoveries. I'd rather not reopen that debate. Fcrary (talk) 04:01, 11 January 2020 (UTC)
 * I don't understand your point. Of course we should distinguish between a failed attempt to land and not making an attempt. The problem is that the graph currently confuses attempts to land with attempts to do an ocean touchdown (or ocean splashdown, in the case of the parachutes). Tercer (talk) 08:40, 11 January 2020 (UTC)

Workshop on captions
As suggested by, we can discuss captions separately from the charts. Given that there is so far no appetite to change the categories of landings documented in the charts, let's discuss how to improve the relevant captions for clarity. The current structure of captions is as follows: I think there is no ambiguity with the top 2 lines: a landing was attempted, either on a ground pad or on a drone ship, and it succeeded or failed. Likewise, the "Parachutes failure" case sounds clear enough: SpaceX tried to land the booster with parachutes, and that failed. That leaves three cases to be discussed and potentially improved:
 * Ocean failure
 * Ocean touchdown
 * No attempt

The "No attempt" case means that SpaceX let the booster fly back to Earth without attempting a controlled landing. This is usually because the mission requirements precluded a controlled re-entry (must burn all fuel for a heavy satellite), although there may be other reasons in some rare cases (e.g. drone ship can't sustain high seas). I think the caption is clear enough as is: SpaceX made "no attempt" to land the booster; it's not the role of the caption to explain why.

The "Ocean failure / Ocean touchdown" pair may require some explanations. The meaning of the current captions is that SpaceX tried to perform a soft vertical touchdown on the ocean surface, and either succeeded or failed. Thte technical definition of "touchdown" is that the booster reached zero vertical velocity (z'=0) when touching the ocean (at z=0). In the early development years (2013–2015), the booster could only target a landing zone of several km diameter, so that SpaceX first focused on perfecting the landing burn to reach z'=0, then gradually improved the horizontal precision. Once the boosters with grid fins were capable of sufficiently narrowing down the x/y target, SpaceX floated a drone ship on the booster's path and attempted pin-point landings on the target. Thus, from 2015 onwards, we switch to drone-ship landing attempts, and we get a bunch of spectacular crashes until the first success in April 2016.

Separately, in late 2017 / early 2018, SpaceX performed two controlled ocean touchdowns without sending a drone ship. This is the time when they were getting rid of pre-block 5 boosters that they had decided were not worth trying to fly a third time (see List of Falcon 9 first-stage boosters for details). Such boosters were usually assigned to high-energy missions in which they burned all their fuel and were allowed to fall back to the ocean uncontrolled, but on some missions they had enough margin for a controlled re-entry, performed some tests of the flight envelope, and just left the boosters "buried at sea". Those are still "ocean touchdowns" for our purposes, albeit for a different reason than in the early years.

Do we need to alter the "ocean" captions for clarity? I don't think so: we can convey a better understanding of what they mean through footnotes briefly explaining what happened as outlined above. For example, "Ocean failure" could be explained with "Controlled descent; booster did not achieve a soft touchdown", and "Ocean touchdown" with "Controlled descent and soft vertical touchdown on the ocean surface". Again, I don't think captions should delve into the reasons for attempting a touchdown, just state what happened. So my proposal is:

Comments welcome. — JFG talk 07:27, 11 January 2020 (UTC)
 * Might be a bit long, but what about "Ocean touchdown failure" and "Ocean touchdown success", to follow the style of the other categories. --mfb (talk) 07:34, 11 January 2020 (UTC)
 * That would be an improvement over the current captions, because it would make it clear that these two are a pair. Tercer (talk) 08:34, 11 January 2020 (UTC)
 * This proposal doesn't address any of the problems with the captions. First of all, "Parachutes failure" is not clear at all. SpaceX did not try to land the booster with parachutes, they tried to make an ocean splashdown à la Shuttle SRB. As for the ocean pair, the footnotes don't explain the key point: they were tests, not landing attempts. Also, they don't look like a pair now, although they are. This is confusing. Furthermore, you are calling both the successful reentry test of flight 6 and the failed touchdown test of flight 13 "ocean failure". That's misleading, and brings us to the final problem: failure/success doesn't really apply to tests, as their goal is to develop landing, not perform an ocean touchdown per se. At least in the list they are called "Ocean (controlled)" and "Ocean (uncontrolled)", which avoids this pitfall. Tercer (talk) 08:33, 11 January 2020 (UTC)
 * "Parachutes failure" sounds clear enough: the goal was a splashdown, assisted by parachutes, and it failed. Regarding the ocean touchdown attempts, which indeed can be characterized as system development tests rather than landing attempts, I could suggest "Controlled descent attempt" with footnote "Booster did not achieve a soft touchdown", and "Ocean touchdown" with the footnote "Controlled descent and soft vertical touchdown on the ocean". Thus we don't have to call those tests successes or failures. What do you think? — JFG talk 11:31, 11 January 2020 (UTC)
 * No, it's not clear. There's absolutely zero indication that the goal was a splashdown instead of a landing. My preference would be to refer to the three as "Test (splashdown uncontrolled)", "Test (ocean uncontrolled)" and "Test (ocean controlled)". The "Test" makes it clear that they were not landing attempts, and in brackets we explain the outcome. As for "Controlled descent attempt", that is a bit ambiguous; does descent refer only to the descent, or also to the touchdown? The goal of flight 6 was just the reentry, but the goal of flight 13 was an ocean touchdown. Flight 6 achieved reentry, but failed descent. Flight 13 achieved reentry and descent, but failed touchdown. As for "Ocean touchdown", it doesn't make it clear that the goal was not a landing, causing the confusion with CRS-16. Tercer (talk) 12:29, 11 January 2020 (UTC)
 * Tests can succeed or fail. They come with a list of things the booster is expected to do, and it does these things or not. We capture that in the table and the graph based on information from SpaceX. Where is the problem? mfb (talk) 13:06, 11 January 2020 (UTC)
 * Of course you're right, mfb. If this graph were about the test, we could point out the goal and say success or failure. For example, flights 1 and 2 would be "Reentry failure", flight 6 would be "Reentry success", flights 9 and 10 would be "Touchdown success", flight 13 "Touchdown failure", and so on. The problem is that the graph is about the landings, and when we write success/failure the reader will assume that it refers to the landing, not to whatever the goal of the test was. Currently there's no way for the reader to even know what was the goal of "Parachute failure" or "Ocean failure". Tercer (talk) 14:04, 11 January 2020 (UTC)
 * 'Test (splashdown uncontrolled)' Doesn't tell you it was with parachutes. I don't see an advantage to 'Test (ocean uncontrolled)' compared to 'Ocean test uncontrolled' which is shorter but still possibly too long. Clarifying which are tests and which are landing attempts is potentially an issue. So I suggest the following: crandles (talk) 14:13, 11 January 2020 (UTC)

If people generally prefer 'Ocean test uncontrolled' to 'Ocean test failure' and there is room for that on a single line then I have no major objection. Not sure the parachute note is needed, but can be used to give more info. crandles (talk) 14:13, 11 January 2020 (UTC)
 * I intentionally removed the mention to parachutes. The most important aspect of those tests is that unlike all others they were about splashdowns, not landings, and I think we should emphasize that instead of the parachutes. Independently of that, I don't like the inconsistency failure/controlled. I prefer to use "test uncontrolled"/"test controlled" for all, but I'm also fine with "test failure"/"test success" for all. Tercer (talk) 16:16, 11 January 2020 (UTC)
 * The most important aspect of those tests is that they use a different method, so number one question is, what is that method. Using parachutes answers this no 1 question. crandles (talk) 19:14, 11 January 2020 (UTC)
 * I think the most important aspect is the goal, not the method. But it's ok, I can live with parachutes. Tercer (talk) 19:18, 11 January 2020 (UTC)
 * The goal includes the method to be well-defined. "My goal is to reach New York City" - okay, but do I drive there, walk there, take the airplane? These are very different things. SpaceX tried to work on recovery via parachutes, that is information that shouldn't get lost. mfb (talk) 21:50, 11 January 2020 (UTC)
 * I could live with "Ocean test failure" and "Ocean test success" for symmetry. Footnotes have details. I strongly object to not mentioning parachutes: they were definitely intended to enable booster recovery. That route failed and was quickly abandoned. — JFG talk 17:44, 11 January 2020 (UTC)
 * , I really like the use of explanatory notes. The parachute notes IMO aren't necessary as this seems super freaking obvious that they tried to use parachutes and it failed. The exact details of that are really not important in this graph. precisely which captions we use is less important with the added footnotes. I'd support most/all of the suggestions made, but probably prefer "Ocean test failure" and "Ocean test success" for symmetry. —  Insertcleverphrasehere (or here) (click me!)    18:10, 11 January 2020 (UTC)
 * Perhaps it should be Parachute test failed rather than failure to indicate it is past tense i.e. SpaceX has given up on this approach. I thought it might be useful to keep success and failure for the landing attempts and I think there is room for Ocean test controlled and Ocean test uncontrolled on the same line, if symmetry is desired. This avoids issue of flight 6 CASSIOPE being a successful re-entry test but uncontrolled. But if success and failure are preferred, no problem. An additional possibility is to extend 'No attempt' to 'No attempt nor test'. crandles (talk) 18:38, 11 January 2020 (UTC)
 * Insertcleverphrasehere, we could use the footnote on "Parachutes test failure" to explain that it was a splashdown attempt. In any case, as it is the footnote is false, as the parachutes never deployed. I suggest "Splashdown attempt; booster burned up on reentry before parachutes could deploy." Tercer (talk) 18:50, 11 January 2020 (UTC)
 * , neither is necessary, this detail is irrelevant to the graph.The reader will understand that there was a parachute test that failed. —  Insertcleverphrasehere (or here) (click me!)    19:01, 11 January 2020 (UTC)

A possible solution
I think I have a solution that everybody will be happy with. Why don't we split the graph into a "Landing outcomes" graph, and a "Landing development" graph? In this way we can properly explain what the tests were about, and use the success/failure categories without danger of confusion with the landing attempts. My suggestion is to use categories based on the EDL scheme: Reentry success/failure, Touchdown success/failure, and Landing success/failure. So flights 1 and 2 would be "Reentry failure", flight 6 would be "Reentry success", flights 9 and 10 would be "Touchdown success", flight 13 "Touchdown failure", and so on. Also, we would be able to put information about weird cases such as Hispasat that are currently listed simply as "No attempt". What do you think? crandles, you previously expressed interest in having a graph in the Falcon 9 first-stage landing tests page, maybe the "Landing development" one would be more appropriate there? Tercer (talk) 19:15, 11 January 2020 (UTC)
 * , No. We've got a good solution in the works with the footnotes in the captions. We don't need even more graphs where the one that we have can accurately and succinctly relay all the relevant information. Please just stop. we have a solution, please just leave the poor horse alone! —  Insertcleverphrasehere (or here) (click me!)    19:46, 11 January 2020 (UTC)
 * Will you please stop with the DEADHORSE insults? I stopped using FANCRUFT a long time ago, and I would appreciate some courtesy in return.
 * The matter of fact is that the graph is too complicated. Improving the captions doesn't solve that, it merely improves the captions. Tercer (talk) 20:04, 11 January 2020 (UTC)
 * , At this point you are engaging in tendentious editing. There is an RfC above with unanimous opposition, and yet you still claim that your opinion is 'fact' and therefore everyone else is wrong and something has to be done to appease you. At this point, your behaviour is becoming disruptive as your constant WP:REHASH is wasting a lot of people's time that could be spent better. —  Insertcleverphrasehere (or here) (click me!)    20:29, 11 January 2020 (UTC)
 * Are you ignoring the empirical evidence before your eyes? We have three people commenting here who are not authors of the article: me, Soumyabrata, and Jadebenn. I maintain that the graph is too complicated, Soumyabrata misunderstood it and wants to simplify it, and Jadebenn wouldn't make a statement one way or the other. This means we either have 66% or 100% of the readers saying that the graph is too complicated. And your reaction is to simply insult me? Grow up! Tercer (talk) 20:39, 11 January 2020 (UTC)
 * , Your method of determining consensus is very curious. Your definition of an insult is also curious. Whatever; I don't care. You've worn me out by constantly coming back at everything everyone has said with the same arguments. You've had your say. Many, many times. In multiple discussion threads. It's a blizzard; let it go. —  Insertcleverphrasehere (or here) (click me!)    21:34, 11 January 2020 (UTC)
 * If you remember, the very purpose of started a RfC, and a DRN before that, is to get readers to come here to determine whether the graph was too complicated. Now that it happened, and the result is not what you wanted, you decide to ignore it and insult me instead. Maybe you should thank the readers for the feedback and admit that there was in fact something unclear with your article? Tercer (talk) 22:08, 11 January 2020 (UTC)
 * , What? You are just moving the goalposts. 'Readers' almost never comment in RfC's and an RfC's outcome is certainly not dependent on the editors that comment on them being 'readers' rather than regular editors. anyway, Soumyabrata is a regular editor on spaceflight articles (including other SpaceX articles), and Jadebenn came here from Wikiproject Spaceflight. If you think that the current result of the RfC validates your opinion, you are being delusional and don't understand how consensus works. —  Insertcleverphrasehere (or here) (click me!)    22:16, 11 January 2020 (UTC)
 * I'm not moving the goalposts. This has been my stated goal since the beginning. Let me quote myself: "My impression is that the editors here are very knowledgeable about the details of every aspect of Falcon launches, and lost sight of what is relevant/understandable to the people who are not. I think the best solution is to post it in WP:DRN to get a view from the editors at large."
 * I'm not talking about determining consensus. I'm talking about asking people whether the graph is easy to understand. It's standard practice in physics (and I assume everywhere else) to give your draft paper to someone who is not an author and ask them for feedback. Precisely because everything is clear for the authors, and it is hard to guess how other people will interpret your writing. Jadebenn and Soumyabrata are not authors of this particular article, so they qualify. Tercer (talk) 22:24, 11 January 2020 (UTC)
 * , This is not a standard way of determining consensus for change on Wikipedia, and trying to make it work that way is not going to happen. in RfCs, every editors view is valuable, not just uninvolved editors. Please accept this. —  Insertcleverphrasehere (or here) (click me!)    22:31, 11 January 2020 (UTC)
 * I wasn't trying to determine consensus. I was trying to determine whether the graph is easy to understand. It is not. It is not an opinion, it is a matter of fact which you are wrong about. Please accept this. Tercer (talk) 22:54, 11 January 2020 (UTC)

Please accept that Wikipedia works on finding a consensus among the editors. It does not get into statistically significant surveys of uninvolved readers. That's just not practical. The assumption is that someone with an opinion on how the material is presented will speak up, and their opinions are valid whether they are they original author, a later editor or a reader who has never edited the article. When there is a consensus (a strong majority), that's what we go with. The strong majority of the people who have bothered to comment is that the chart is not too complicated. That's a fact. You can't alter it by disqualifying people's opinions. Fcrary (talk) 00:25, 12 January 2020 (UTC)
 * My hope was that once I had demonstrated that the graph was confusing you would actually change your minds, and so there would be a consensus for doing something about it. How foolish I was. You clearly don't care about making the graph understandable. Tercer (talk) 00:46, 12 January 2020 (UTC)
 * At this point, we're not only in a blizzard, but it's 1 AM. Seriously, great advice to read there. — JFG talk 10:56, 12 January 2020 (UTC)
 * Indeed, there's some great advice there. Even something for you: "The experienced editor: This editor has been around a long time, has made many edits, has no recent blocks, and generally gets along with everyone. In this case you should seriously reexamine your own position, especially if you are a fairly new editor. Work with the lone holdout and try to figure out why you are – or think you are – in a situation that almost never happens. Figure out what is going on. Ask a third party to look into it if needed.". Tercer (talk) 12:58, 12 January 2020 (UTC)
 * Gosh. "By the way, if you are the one who is going up against the many yet you find yourself ignoring the advice that was created for you and instead find yourself studying this section so you can play "gotcha" against the many, you need to rethink your life choices." Nuff said. — JFG talk 14:27, 12 January 2020 (UTC)
 * Yet you're studying that page to play "gotcha" with me. Oh the irony. Of course you'd never consider the possibility that you need some advice yourself. Tercer (talk) 15:27, 12 January 2020 (UTC)
 * Congratulations! Although I did my best to engage with you and accommodate your concerns about this graph, you have now exhausted my supply of WP:AGF. On Wikipedia, nobody is obliged to WP:SATISFY you. Do not expect further replies and overtures from me. — JFG talk 18:23, 12 January 2020 (UTC)
 * , There are only so many times that you can play a reverse Uno card before you just start looking silly. —  Insertcleverphrasehere (or here) (click me!)    20:31, 12 January 2020 (UTC)
 * WP:PA Tercer (talk) 20:44, 12 January 2020 (UTC)


 * Tercer, I'm not sure why you say only three people in this discussion are not "authors" of this article, or why you call this empirical evidence. I just checked: I've edited this article exactly three times in the past year, and one of those edits was to correct a typo. What is your empirical evidence that I'm an "author" and somehow biased? Fcrary (talk) 00:14, 12 January 2020 (UTC)
 * I didn't check that, I just assumed that everyone who was discussing this on the talk page was an author. I'm not claiming that authors are biased, I'm just saying that if you want to know whether your paper is clear, you have to give it to somebody else to read. This is standard practice in physics. Tercer (talk) 00:23, 12 January 2020 (UTC)
 * I guess that was my point. You assumed that everyone who was discussing this on the talk page was an author, without evidence, and then you stated that as an incontrovertible fact. And you go on to offer opinions about how a professional scientist would behave. I find that inconsistent. As far as finding other people to read your papers before submitting them, sure. Actually, I'm more likely to present the results at a conference, but that's a detail. But having other people look at the work is what Wikipedia is all about. Anyone can read and comment (and edit.) But if it's just one person who has a problem, the comments aren't going to be taken too seriously and the edits will get reverted. That's where consensus comes in. Fcrary (talk) 00:55, 12 January 2020 (UTC)
 * If you don't consider yourself an author, fine, I'm sorry for stating that. But please stop pretending that I'm the only one that has a problem with the graph. Tercer (talk) 07:52, 12 January 2020 (UTC)
 * IMO, yes I do think it can be overwhelming to see all of those labels on graph. And if there was a way to convey what we have now without losing any info I would support it. OkayKenji (talk page) 07:59, 12 January 2020 (UTC)
 * , that's the rub isn't it? It may be a complex, but so was the development of the first orbital booster that can land and be recovered. If we simplify the graph in the ways Tercer suggests, we lose info. IMO the graph is as complex as it needs to be to convey the information that it needs to convey. Still there are ways to improve it, such as the explanatory footnotes and captions that are being worked through in the section above, which are fundamentally good ideas I think. —  Insertcleverphrasehere (or here) (click me!)    08:39, 12 January 2020 (UTC)
 * , well yes, that's precisely why I'm suggesting to add a "Landing development" graph. In this way we would gain, not lose, information. Tercer (talk) 12:42, 12 January 2020 (UTC)
 * Everyone rejected your proposals to change the graph, and now you think everyone will be happy with changing the graphs? Are you serious? mfb (talk) 21:52, 11 January 2020 (UTC)
 * Yes, I'm serious. What is abundantly clear is that everyone here cares about the details of the landing tests. Fine. I'm proposing to keep them and present them properly. Currently there's only a messy summary of them. In a dedicated graph you can actually show how the development went. Tercer (talk) 22:14, 11 January 2020 (UTC)
 * , Don't you see that people can simultaneously believe that the graph is complicated, and that it should also remain the way it is? Sometimes something just has to be as complex as it has to be and simplifying it just results in a worse product. I've already said this a couple times to you on this page (in different words), yet you seem unable to grasp the concept and act like anyone who agrees that the graph is complex automatically supports change. That's just not the case, empirically shown in the results of the RfC. If the current graph is complex... oh well? It's the best we could come up with to show what we wanted to show, and nobody has come up with anything better. You gave it your best, the !votes are in, your solutions weren't considered better than the current graph. There is a certain point where you cant keep going "Well what about...?" and expect people to continue to humour you as you move on to the next proposal before the ink is dry from the !votes of the previous one. —  Insertcleverphrasehere (or here) (click me!)    10:41, 12 January 2020 (UTC)
 * Holding this position is saying that you don't care whether the reader understands the graph. I don't find this acceptable. Please keep in mind that you don't write the article for yourself to read, but for other people. Tercer (talk) 13:18, 12 January 2020 (UTC)
 * , are you even listening? “I don’t find this acceptable.” Too bad, you’re outvoted, that’s consensus. If you continue to argue in the face of obvious SNOW opposition then I’ll have to take this to WP:ANI as a clear case of disruptive editing. —  Insertcleverphrasehere (or here) (click me!)    20:04, 12 January 2020 (UTC)
 * And now you resort to threats?! Please do take it to WP:ANI, you'll get a reality shock. Tercer (talk) 20:17, 12 January 2020 (UTC)
 * , If you want to interpret it that way, fine. I don't want to go to ANI, I hate that place, but you aren't leaving us with any good options. Your exchange with JFG above is illuminating. Please... when everyone else is telling you something, and you vehemently believe something else, that might be an indication that you need to A) reconsider your position, or B) let it go. Given your previous comments, A seems unlikely, so I'd suggest you take option B. —  Insertcleverphrasehere (or here) (click me!)    20:24, 12 January 2020 (UTC)
 * As was the case with JFG, you always have advice for me, but don't consider the possibility of applying that advice to yourself. Maybe you could reconsider your position? Since my first edit you have adamantly refused to do anything to simplify the graph. I really don't understand how you find defensible your position of (paraphrasing) "the graph is complicated, but we shouldn't do anything about it". Tercer (talk) 20:31, 12 January 2020 (UTC)
 * , Um... if you'll read back, I tried to work with you directly to improve the captions, made suggestions and worked out which options were too long. You seem dissatisfied with this compromise solution, but that discussion seems to be bearing fruit.
 * There are two opposing issues here: 1) too many categories in the graph (which you seem to consider a major issue and many others, including me, consider a minor issue), and 2) having the chart show the progression of the development of different landing tests and methods (which most people who !voted in the RfC consider to be of major importance and you consider of minor importance). Regarding your 'simplification', which really boils down to removing the progression of tests from the graph; that is less important to everyone else than having all the info in the graph. If others agreed that simplifying the graph was more important that displaying that info, I would bow to consensus. That isn't the case. So yes, "the graph is complicated, but we shouldn't do anything about it" (at least with regards to removing categories from the graph). —  Insertcleverphrasehere (or here) (click me!)    20:49, 12 January 2020 (UTC)
 * , Also, I've been here before, I previously thought it might be a good idea to change this graph to a 'booster recovery' graph, but that comes with it's own issues, and there was clear opposition to that change back then as well. Since then, I've come around to the thinking of others that the progression of test is more important than having a simple graph with just recovery success/failure. One of the main reasons for this is that not all the boosters that landed were recovered, one of the falcon heavy cores landed successfully then was destroyed at sea. Also, for SpaceX it seems that recovery was not always an important goal; a lot of pre-block 5 cores were intentionally sent in at dangerously fast speeds to do ocean tests; after they had the block 5 cores they didn't need these boosters, even if they could have recovered them. Defining these as 'ocean test success' is accurate, but they also were intentionally declining to recover those boosters, so 'no attempt' sort of fits from a recovery standpoint for these boosters. But to muddy things further, from a recovery standpoint 'no attempt' does not really fit for the early water tests, where they actually did intend to try to drag them back to shore (but they exploded on tip over). On top of that, they DID actually drag one of the later water test boosters back to shore, even though they hadn't intended to recover that one and had expected it to be destroyed at sea. There was also the ground pad failure booster, which they dragged back to port and was partially recovered (at least the grid fins were).
 * All in all, there's no real way to capture all of this in a 'Booster recovery' graph, not without being factually inaccurate with some of the individual data points. So ultimately, the recovery graph option does not really work and I figured this out quite a while back. The option left is to just display all 7 categories in the 'landing' graph and live with it. Links and footnotes in the captions will help deal with the readability of the graph though. —  Insertcleverphrasehere (or here) (click me!)    21:13, 12 January 2020 (UTC)
 * I'm sorry, I don't think improving the captions is any sort of compromise. It doesn't simplify the graph at all, on the contrary, it precludes simplification. I always said that this was a futile endeavour, but I engaged with you on the subject because you asked.
 * I offered several different compromises: the three options in the RfC, that are all more complex than my original edit, and my current suggestion, of splitting the graph into "Landing outcomes" and "Landing development". I don't see what is the problem with splitting the graph, the information you all care about is still being displayed.
 * My suggestion is having a "Landing outcomes" graph, instead of a "Booster recovery" graph. As such it avoids the pitfalls you mention. Not that I think they're so problematic: flights 46 and 48 are "no attempt" with regards to recovery and the lost centre core is "recovery failure", leaving the other cases as the unsatisfactory ones. A classification can be done by strictly defining "recovery success" as recovery of an intact (modulo salt water) booster when that was the intention. This would put the early water tests as "recovery failure" (even though recovery wasn't really the point), the ground pad failure as "recovery failure", and the accidentally recovered booster as a truly unsatisfactory "no attempt" (I didn't know about this, which flight is it?).
 * There's no avoiding factual inaccuracy with such a complex dataset, though. Remember that the current graph displays flight 6 as "ocean failure", even though it was actually a successful reentry test. Tercer (talk) 22:24, 12 January 2020 (UTC)
 * , That sounds like an absolute mess honestly, with at least as much complexity and more ambiguity with regards to showing what actually what happened when the boosters landed. 'Landing outcomes' seems ambiguous to me in meaning, as you can define this as the immediate landing or the outcome days later, which is confusing and does not match the definitions used in the list (likely this is a deal-breaker). I don't expect I'll support but feel free to mock it up. —  Insertcleverphrasehere (or here) (click me!)    22:37, 12 January 2020 (UTC)
 * "Landing outcomes" would be the option 3 from the RfC. It matches the definitions from the list by design. Tercer (talk) 22:44, 12 January 2020 (UTC)
 * , The other graph would still need to be mocked up. Also, it does not match, obviously. The tests are all defined as 'no attempt', where the landing in the list shows the tests. —  Insertcleverphrasehere (or here) (click me!)    22:50, 12 January 2020 (UTC)
 * Ah yes, it matches apart from the part that obviously doesn't match. I don't see much of a problem with that, we can just modify the descriptions there to include the word "experiment" or something to make it clear that they were not landing attempts, and thus go into "no attempt". As for the other mockup, I hope you'll understand that I'm not eager to do work that you're telling me in advance that you will reject. Tercer (talk) 23:08, 12 January 2020 (UTC)
 * , Well, fair enough. It appears that this section on 'a possible solution' is simply Option 2 (which gained no support in the RfC, with the addition of a second graph (which hasn't been seen and IMO is probably not likely to satisfy a majority), I don't think this is something that "everybody will be happy with" so I agree that it probably isn't an avenue for further consideration. Sorry that the outcomes here aren't to your liking, I have some experience being in similar situations myself (one particular RM at Genesis creation narrative comes to mind vividly). —  Insertcleverphrasehere (or here) (click me!)    23:30, 12 January 2020 (UTC)
 * It's option 3, but no matter. Don't you have any idea about how to simplify the graph? All you're doing is shooting down everything I come up with. Tercer (talk) 23:43, 12 January 2020 (UTC)
 * , I do not. If I did, I would suggest it. —  Insertcleverphrasehere (or here) (click me!)    00:01, 13 January 2020 (UTC)
 * , well, actually I did think of one thing. We could potentially get all the 'test' landings, and put them in different shades of a single colour, this might make the graph easier to parse. The main issue here is that it is quite difficult to do this while also dealing with colorblindness concerns. We have to consider red-blind, green-blind, and blue-blind and make sure that all the colours can be differentiated, which limits our colour choices considerably. I will try to see if I can do something regarding this. —  Insertcleverphrasehere (or here) (click me!)    00:13, 13 January 2020 (UTC)
 * , See below. —  Insertcleverphrasehere (or here) (click me!)    01:02, 13 January 2020 (UTC)
 * , I dug through the list and the flight I was thinking about was flight 48, though SpaceX did actually decline to recover the booster despite it not exploding and technically being recoverable (I was remembering this incorrectly as they didn't drag that one back to shore, only the failed ground pad one). Probably not an issue one way or the other. —  Insertcleverphrasehere (or here) (click me!)    22:41, 12 January 2020 (UTC)
 * Ah, ok. Thanks for digging. Tercer (talk) 22:44, 12 January 2020 (UTC)
 * Re: In a dedicated graph you can actually show how the development went. That's a level of detail relevant to the dedicated article Falcon 9 first-stage landing tests. Try it there. — JFG talk 10:46, 12 January 2020 (UTC)
 * Indeed, that's my opinion as well. I've asked crandles about it, but they haven't answered yet. Hopefully winter is not so bad wherever they are and they have something better to do on the weekend. Tercer (talk) 13:04, 12 January 2020 (UTC)
 * At some point landing development graph on that page may have to diverge from the one here. I am more inclined to sort out preferred captions here and if, with footnotes, that shows all we want to show then the only change might be to stop showing 2020 onwards landings on the landing tests page. However, if you want to experiment with two graphs on the landing tests talk page to demonstrate that it works better, then I can't really stop you. crandles (talk) 14:32, 12 January 2020 (UTC)
 * That's a polite way of saying no. Ok, nevermind then. Tercer (talk) 15:07, 12 January 2020 (UTC)

Compromise?

 * . Perhaps we could use colour to differentiate the test landings from the others and make the graph easier to parse? —  Insertcleverphrasehere (or here) (click me!)    00:43, 13 January 2020 (UTC)

 Current 

 Proposed 

Here I have used gray shading for each of the tests and 'No Attempt' and used colours for the Ground pad and drone ship landings. I have also used some of the suggestions in the above captions workshop with footnotes. Note that I have had to use "Parachute" rather than "Parachutes" because with the note it does not fit on the same line. There is still more room to refine the captions and footnotes of course. —  Insertcleverphrasehere (or here) (click me!)    00:43, 13 January 2020 (UTC)

And yes, this works for all the colourblindness variants. Though I've shifted FireBrick and SeaGreen to DarkRed and ForestGreen for better contrast (also to match the colours used in the adjacent launch graph). —  Insertcleverphrasehere (or here) (click me!)    00:46, 13 January 2020 (UTC)
 * That's... surprisingly good, I must admit. I doesn't actually simplify the graph, but it creates a clear visual separation between the landing attempts and the tests. Coupled with the new captions, I think it won't confuse readers anymore. I'm fine with it. Tercer (talk) 09:00, 13 January 2020 (UTC)


 * Pinging RFC !voters and previous discussion participants:, , , , , , , , , . Something that we can all live with? —  Insertcleverphrasehere (or here)  (click me!)    12:16, 13 January 2020 (UTC)
 * we all lost infinite hours to end at the beginning + some footnotes & color changing? (yes, i was to mantein actual infograph.....but it doesn't matter). next time be more quick to come back to beginning. in any case APPROVED--Dwalin (talk) 12:55, 13 January 2020 (UTC)
 * No objection crandles (talk) 13:14, 13 January 2020 (UTC)
 * Neither any objections from me. Gial Ackbar (talk) 17:30, 13 January 2020 (UTC)
 * Yeah, no objections. OkayKenji (talk page) 20:29, 13 January 2020 (UTC)


 * Looks good (and I prefer this over the version below where "no attempt" is too similar to the background). --mfb (talk) 23:17, 13 January 2020 (UTC)

New colors take 2

 * It's acceptable to denote the various test scenarios by a series of shades of grey; however I'd tweak the colors as above, so that "no attempt" flights are clearly distinguished from both development tests and landing attempts. I also copy-edited the footnotes for brevity, and restored the "Ocean touchdown" outcome which I find clearer than "Ocean test success". In that spirit, we could also change "Ground-pad success" and "Drone-ship success" to "Ground-pad landing" and "Drone-ship landing", but I'm not sure whether that's clearer. Please comment. — JFG talk 18:21, 13 January 2020 (UTC)
 * I find both Insertcleverphrasehere's colours and captions easier to understand. In any case, captions can be discussed independently of the colours. Tercer (talk) 19:04, 13 January 2020 (UTC)
 * , Definitely further room to decide the captions. I do prefer the light grey to that cream colour in the second graph. At least on my monitor it doesn't give good contrast with the white background. Also, the FireBrick and ForestGreen don't play well together for colourblindness (which is why I used DarkRed). —  Insertcleverphrasehere (or here)  (click me!)    00:49, 14 January 2020 (UTC)
 * I'll be honest. This discussion is going in circles over trying to fix something that isn't broken. The graphs were fine but looking at what has proposed is the best way to go. Giving us a bit more info, changing colours sloghtly to match it but not changing the way the graph works and making it look confusing by over simplifying. The colours and layout  showed is the better one in my opinion and I fully support a change to that, just so this discussion can come to an end. - AndrewRG10 (talk) 7:32 14 January 2020 (UTC)

New colors take 3

 * I have taken into account general preference for 's color scheme, so I'll bow to that and just suggest a slight tweak, replacing the "Gold" color with "Goldenrod" for ground-pad failure, because gold looks too bright compared to adjacent colors. Please check for color-blind accessibility. Separately, I'm testing a more legible color for "no attempt", given remarks that "Linen" was not distinctive enough from the background white. Comments welcome on both variants. I have also made the charts slightly larger to avoid text collisions (bonus: "Parachutes" can be plural again). — JFG talk 08:54, 15 January 2020 (UTC)


 * Goldenrod adjustment


 * Goldenrod + distinguish "no attempt" from tests

Final comments, anyone? — JFG talk 08:54, 15 January 2020 (UTC)
 * , Yeah the goldenrod looks fine in Corblis. I'm personally fine with the new colour for 'No attempt'. —  Insertcleverphrasehere (or here) (click me!)    09:09, 15 January 2020 (UTC)
 * I think goldenrod is an improvement over Insertcleverphrasehere's version, but I still think their colour for "No attempt" is better. The footnote for the parachute category is a bit misleading, as it suggests that parachutes were actually used in the test. I'd replace that with "Non-propulsive reentry attempt; booster destroyed before parachutes could deploy". Tercer (talk) 09:18, 15 January 2020 (UTC)
 * "Non-propulsive reentry attempt" You can't use parachutes and propulsion at the same time for long, so I think that is fairly close to being redundant. 'Parachutes to control decent velocity only' conveys more information as this is then clearly aiming for a splashdown somewhere rather than a specific location as you are saying the parachutes are not steerable. I would prefer to add '; method failed before deployment' which gives more info than just this particular booster being destroyed before parachutes could deploy. Is 'Parachutes to control decent velocity only; method failed before deployment' too long or other issues with this? crandles (talk) 12:36, 15 January 2020 (UTC)
 * The idea back them was to do passive reentry, and afterwards use parachutes for a soft splashdown. You can't use parachutes for reentry, or at least nobody ever tried. And because passive reentry failed, SpaceX decided to do propulsive reentry. This is logically independent of doing propulsive landing, even though SpaceX simultaneously went for propulsive reentry and propulsive landing. It's perfectly possible to do propulsive reentry and then use parachutes for a splashdown, even though this never happened. Of course you wouldn't use propulsion and parachutes simultaneously, that would be silly. Tercer (talk) 13:36, 15 January 2020 (UTC)
 * So, if you want to be clear about this, isn't something like 'Passive reentry failed before parachute deployment' clearer and simpler than your version?crandles (talk) 14:38, 15 January 2020 (UTC)
 * Indeed it is, I would prefer your footnote. Tercer (talk) 14:46, 15 January 2020 (UTC)
 * I'd support the reworded footnote: "Passive reentry failed before parachute deployment". Was it a single parachute or several? — JFG talk 15:16, 16 January 2020 (UTC)
 * Several. Tercer (talk) 15:33, 16 January 2020 (UTC)
 * Hmm. So is it parachutes deployment or parachute deployments or parachutes deployments or apostrophe(s) needed or something else? crandles (talk) 19:44, 16 January 2020 (UTC)
 * I'm sorry, I can't help with that, not a native speaker (or a good writer, for that matter). Tercer (talk) 19:59, 16 January 2020 (UTC)

Should the captions be reversed in vertical direction? This would put early tests near the top so more in line with usage in the graph. Also it puts the parachutes note first then the two ocean splashdown notes adjacent rather than the parachute note being in the middle. crandles (talk) 12:36, 15 January 2020 (UTC)
 * I think that's a good idea. Tercer (talk) 13:36, 15 January 2020 (UTC)
 * Reversing the order of captions would also reverse the order of colors in the graph. Better keep them as is. — JFG talk 15:17, 16 January 2020 (UTC)
 * That's a good point, better leave it as it is. Tercer (talk) 15:33, 16 January 2020 (UTC)

Trying it as I think it makes more sense: order should go from failures to neutral no attempt to sucessful tests to successful landings.

Hmm, didn't like look of it with AntiqueWhite, even if I think order makes more sense. crandles (talk) 16:26, 16 January 2020 (UTC)
 * Please don't, it looks hideous. Why can't we just use 's version and be done with this discussion? Tercer (talk) 19:59, 16 January 2020 (UTC)
 * Gone back to Gainsboro grey. Also tried tweaking footnotes. crandles (talk) 22:18, 16 January 2020 (UTC)
 * What? No, it's touchdown, not splashdown. Splashdown is easy, has been done by the shuttle SRBs and Apollo capsules. Touchdown is the noteworthy aspect.
 * Also, the issue with "no attempt" is that it is the least relevant category. As such an unremarkable colour like Gainsboro grey is perfect. Also it's important to leave their bars on top of the other bars, in order to not affect their distribution. Otherwise you're making it more difficult for people to see the most relevant information. Tercer (talk) 07:01, 17 January 2020 (UTC)
 * , I also agree that no attempt should stay at the top so that it doesn’t get put underneath the landing successes/failures —  Insertcleverphrasehere (or here) (click me!)    10:28, 17 January 2020 (UTC)

'No test nor landing' rather than 'No attempt'? (Then perhaps there is less need to distinguish by colour?) crandles (talk) 12:36, 15 January 2020 (UTC)
 * No. A lot simpler to just say no attempt. I think the golden colour with the grey for no attempt as proposed is just right, if no one has strong objections I think it's fine to put it on the main page. - AndrewRG10 (talk)) 0:25, 16 January 2020 (UTC)

OK, lets go for 'Goldenrod adjustment' first graph under 'New colors take 3' changing only the parachute footnote to 'Passive reentry failed before parachutes deployments' crandles (talk) 11:26, 17 January 2020 (UTC)
 * Great, I'll do the edit and close the RfC then. I asked an Australian friend of mine, and he told me that the correct sentence is 'Passive reentry failed before parachute deployment', even though there were several parachutes being deployed. The point is that "parachute" in this sentence is an adjective, not a noun, so you don't make it plural. You would if the sentence were 'Passive reentry failed before deployment of the parachutes', for example, as in this case "parachute" is a noun. Deployment in any case was only one (well, none, but one was planned). Tercer (talk) 12:31, 17 January 2020 (UTC)


 * Drogue chute deployment then parachute deployment so two deployments planned? crandles (talk) 13:42, 17 January 2020 (UTC)
 * One can define "deployment" to be the whole process, but as you wish. Are you sure they were going to use drogues, though? Tercer (talk) 13:56, 17 January 2020 (UTC)
 * No I am not sure, which is why it was a question and I didn't edit article. I'm ok with defining it as whole process and leaving it but if someone does know or ideally has a ref...crandles (talk) 16:02, 17 January 2020 (UTC)
 * I don't know it either, but I would bet a lot that they had a drogue packed in, just because every single spacecraft uses them for landing or splashdowns: Soyuz, Shuttle SRBs, Starliner, Dragon... As you wish, though, I'm fine with either deployment or deployments.

Launch statistics by customer
As there are going to be a lot of Starlink launches, I think that it would be usefull a chart that breaks down the launches by primary customer: NASA, USAF, Commercial, SpaceX.

Further, the Crew Dragon test launches should probably be listed as SpaceX as the customer, as those are not operational launches for NASA.

Also, a gray area is Spacecom AMOS-17 as this is not exactly a customer launch, is more of a SpaceX launch in compensation for the lost satellite. — Preceding unsigned comment added by 173.184.78.213 (talk) 01:35, 12 October 2019 (UTC)
 * Amos-17 was part of activities for commercial launches. The Crew Dragon demo missions are part of NASA contracts (same for Dragon 1). How would we treat launches that don't have a clear primary customer? Flight 55 had Iridium and NASA as customers. --mfb (talk) 01:42, 12 October 2019 (UTC)

I think it's an interesting insight to see how the launch market evolves over time, that's why I proposed the chart, and I don't think there is mutch oposition to that.

Second part, how some of the launches are going to be charted, that is controversial, and I can understand why.

If Amos-17 is going to be listed as "commercial" then one contract is going to be counted twice. That is problematic. It could be counted as a goodwill launch by SpaceX, but that is also problematic. — Preceding unsigned comment added by 173.184.78.213 (talk) 13:14, 12 October 2019 (UTC)
 * We could leave out Amos-6 because its rocket never tried to fly. But ultimately SpaceX was handling two satellites as part of one contract, so what is wrong with listing both? --mfb (talk) 15:33, 12 October 2019 (UTC)
 * It would also be interesting that under NASA launches we have three separate colors, CRS, CC, and other NASA satellites. OkayKenji (talk page) 19:11, 12 October 2019 (UTC) nevermindOkayKenji (talk page) 19:14, 12 October 2019 (UTC)

Let's keep it simple here. If we'd like to analyse the launch market, there's an article for that. — JFG talk 13:11, 13 October 2019 (UTC)
 * This is not about an overall market analysis. This is showing how SpaceX gets both government and commercial contracts and starts launching their own stuff now, something unique among launch providers as far as I know. --mfb (talk) 03:44, 14 October 2019 (UTC)

Sample

 * Perhaps if we to add this we have to define what we mean by "SpaceX", "NASA", "USAF", and "Commercial". What do we do if the customer is a government agency or another military-related mission? or like NOAA? Zuma? (sorry I don't mean to sound condescending). The table might look something like this. Please 'do not add this to the main page yet...not sure if the colors are right or if the information is correct, its just an sample of how the table might look like... OkayKenji (talk page) 04:30, 14 October 2019 (UTC)
 * Maybe we can put all non-NASA US government agencies together. How to call it? Your table has much more entries than launches, I fixed 2010 and 2012 but it needs a new counting. --mfb (talk) 15:07, 14 October 2019 (UTC)
 * Its because some launches had more the one customer... OkayKenji (talk page) 15:23, 14 October 2019 (UTC)
 * The suggestion was to go by primary customer. If you list payloads instead of launches then the graph will explode from the ride-share launches, and listing a small cubesat equivalently to a big GEO communications satellite doesn't feel right ″to me. --mfb (talk) 02:05, 15 October 2019 (UTC)
 * Thank you! Sorry I missed that. Still thinking about the name though. I’ll re-count soon. (also we may need to check if the colors are color-blind accessible) OkayKenji (talk page) 05:09, 15 October 2019 (UTC)
 * Coblis check: Red-blind makes blue/purple indistinguishable, red/green become very similar. Green-blind struggles with the same combinations, but not as bad. It makes the yellow very similar to the background, however. Blue-blind has the yellow/background problem as well. We can reuse a color scheme from existing graphs. Take the colors from booster landings for example: Black for SpaceX, green for NASA, blue for commercial, yellow for USAF, grey for other. --mfb (talk) 05:47, 15 October 2019 (UTC)
 * Strongly Oppose It's already complicated enough having Internal missions and customer missions as most people who look at the graph just want to see how many flights SpaceX has done that year. They don't want to count up all the different sections to find out mission numbers. If they want to know who bought the flight can scroll down and do the research below. --AndrewRG10 (talk) 19:36, 15 October 2019 (UTC)
 * If you just want to see the number, you look at the launch outcomes graph (only future missions are split between internal and external). crandles (talk) 10:06, 16 October 2019 (UTC)
 * This is proposed as additional graph. Nothing would go away. --mfb (talk) 15:36, 16 October 2019 (UTC)
 * I think this is a great idea. The colors should change though. — Preceding unsigned comment added by 72.24.143.230 (talk) 02:09, 20 December 2019 (UTC)

It looks great, you can already tell which customers scaled back lately — Preceding unsigned comment added by 2601:CA:8102:6B60:ED3B:1B5F:544E:227B (talk) 00:39, 5 November 2019 (UTC)
 * I like the idea of this graph. The colours are extremely jarring though, I'd consider using some of the colours from the other graphs which we know work well together but aren't so harsh (I don't think this would be suitable for colourblindness reasons anyway). —  Insertcleverphrasehere (or here) (click me!)    05:37, 6 November 2019 (UTC)


 * do we have an update on this graph with a re-count and new colours? Maybe try using some combination of the colours in the Booster landing's graph, since we know those work for colourblindness? I still like the idea of this graph. —  Insertcleverphrasehere (or here) (click me!)    21:23, 12 January 2020 (UTC)
 * I recounted them, (see the table). I have not converted it into the graph yet. I think the different customers are going to be "SpaceX", "NASA", "Commercial", "USAF", and "NRO". Perhaps there is a way to combine USAF and NRO customers? OkayKenji (talk page) 02:50, 13 January 2020 (UTC)
 * , NRO=National Reconnaissance Office? Perhaps put the one NRO launch in 'Other' and add a footnote with the various agencies and customers that fit this category? —  Insertcleverphrasehere (or here) (click me!)    03:13, 13 January 2020 (UTC)

Sample 2

 * , how does this look? For the other graph "Other" was included because I listed all the customers rather then just the main customer, this graph just has the main customers. (here is the graph as a table on my sandbox page). Combined "USAF" and "NRO" (National Recnsece Office" as one since they are both US Gov launches. OkayKenji (talk page) 03:40, 13 January 2020 (UTC)
 * I'd call the fourth group "Military" instead of "USAF/NRO", and I'd add things like Zuma and the Luxembourg government's spy satellite. That would match the classification we use in the navbox Template:SpaceX rocket launches. In that spirit, we could add a "Scientific" category as well, thus distinguishing NASA's ISS-related missions from their (and others') scientific satellites. — JFG talk 07:50, 13 January 2020 (UTC)
 * Go it, generating chart now. (to see how it would look like) OkayKenji (talk page) 20:48, 13 January 2020 (UTC)

Sample 3
While you were making that I was messing around in with the colours in Corblis a bit, and came up with this colour scheme. I tried to match the NASA Blue and Air Force blue (at least as close as web colours get) to their official colours (air force blue itself is a rather notable colour). Not necessary, but kinda cool IMO. It works for Colour blindness, but, I checked yours too and it also works well (while with blue-blind the colours are similar, the shades are distinct). Since I went through all the work I thought I'd share what I came up with anyway. I'll leave the colour scheme up to which one you and others prefer. The Chart itself looks good, I like your approach to the USAF/NRO; solves the issue nicely. —  Insertcleverphrasehere (or here) (click me!)    03:54, 13 January 2020 (UTC)
 * I actually like this color scheme better since its unique to this chart. OkayKenji (talk page) 04:09, 13 January 2020 (UTC)

JFG's suggestions, Insertcleverphrasehere's colors

 * Here are the proposed graphs. might be able to change the "LightSeaGreen" for the "Scientific" with another color that makes more sense? OkayKenji (talk page) 22:36, 13 January 2020 (UTC)
 * I think both versions are clear, but OkayKenji's colours are more beautiful. As for the categories, I think JFG's make more sense. Tercer (talk) 09:29, 14 January 2020 (UTC)
 * Where do scientific NASA missions go in this scheme? --mfb (talk) 21:08, 14 January 2020 (UTC)
 * I'd place them under "Scientific", and rename the Dragon missions "NASA/ISS" or "ISS transport". — JFG talk 08:30, 15 January 2020 (UTC)
 * Fixed. OkayKenji (talk page) 23:11, 15 January 2020 (UTC)
 * Sounds good. Can we give Scientific and Military more contrast? Maybe .... solid Green for scientific? --mfb (talk) 23:15, 15 January 2020 (UTC)
 * , I... do not like this. This is editorialising. We shouldn’t be categorising like this where we have to pick categories that aren’t very clearly defined (what about military satellites that are also scientific?) I preferred the version that just listed the client (who bought the launch). —  Insertcleverphrasehere (or here) (click me!)    10:32, 17 January 2020 (UTC)
 * I don't mind either way. We could also split "NASA (ISS transport)" and "NASA (other)" if people prefer that. Oh, and we should think about crewed launches. A separate category for them, independent of the customer? --mfb (talk) 22:53, 17 January 2020 (UTC)

Landing Categories and colours
Should we edit the categories in the list so that they match the captions? Tercer (talk) 16:13, 17 January 2020 (UTC)
 * One extra line for Parachute Test failure with 2 refs and a footnote is no great disaster for consistency with graph. Ocean test failure and success may fit without extra lines but possibly seems less informational in space available with footnotes a long way away. Not sure if that justifies different text of uncontrolled and controlled. crandles (talk) 16:54, 17 January 2020 (UTC)
 * There is also issue of colours. Do we want as many as 8 colours for the landing column? Black as one of the colours would necessitate white text. At present we use green for success whether on drone ship or land. So I don't think we want to go as high as 8. If it isn't exact match, how many colours should there be? crandles (talk) 16:54, 17 January 2020 (UTC)
 * No. It's perfectly fine as is, no need to intentionally make it more complicated. The graph is there for that reason, landing categoies is just for simply saying what worked and what didn't. - AndrewRG10 (User talk:AndrewRG10) 21:34, 17 January 2020 (UTC)
 * I don't think it makes sense to match the colours, they perform very different functions in the table and in the graph. They didn't match before the change, by the way. I think we should leave the colours as they are (they're good), and change the description to match the captions of the graph, otherwise the reader might wonder whether we're talking about different things. As for the footnotes being far away, I don't think that's a problem, as the list has a long description which is much more informative anyway. Tercer (talk) 16:58, 18 January 2020 (UTC)
 * , I generally would support changing the test in the column to match the graph captions. —  Insertcleverphrasehere (or here) (click me!)    22:06, 18 January 2020 (UTC)

Colors for reused
I think reused boosters would look nicer in green, and would stand out more clearly. I think switching 1.2 reused color with B5 new color would look better and be more clear. 4.35.246.19 (talk) 23:24, 18 January 2020 (UTC)
 * What do you want to make green? Reuse became standard, in the table that would add too much color I think. It would also be confusing to use the same color used for "success" to the right. --mfb (talk) 00:35, 19 January 2020 (UTC)

went boom
went boom for the IFA is the official designation? isn't better to say: "no attempt" for the landing?--Dwalin (talk) 20:17, 19 January 2020 (UTC)
 * Yes. We could note in the description that it went boom as expected. OkayKenji (talk page) 20:23, 19 January 2020 (UTC)
 * ok, i was surprised. --Dwalin (talk) 20:29, 19 January 2020 (UTC)
 * I believe the technical term is Rapid Scheduled Disassembly ;p Tercer (talk) 20:36, 19 January 2020 (UTC)
 * , I'm sorry... What? No. "No Attempt" is what we should use. 'went boom' does NOT describe the 'landing' of this booster. —  Insertcleverphrasehere (or here) (click me!)    20:57, 19 January 2020 (UTC)
 * Rather, we should use 'Precuded' as we did with Launch 19. —  Insertcleverphrasehere (or here) (click me!)    20:59, 19 January 2020 (UTC)
 * I never said that "went boom" should be used. I was agreeing that "No Attempt" should be used. I was saying whereit says "An atmospheric test of the Dragon 2 abort system at Max Q.[493] The spacecraft deployed parachutes and splashdown in the ocean. The test was previously slated to be accomplished with the SpX-DM1 capsule;[494] but that test article exploded during a ground test of SuperDraco engines on April 20, 2019.[418] The abort test used the capsule originally intended for the first crewed flight.[495]" we can say the booster was lost. OkayKenji (talk page) 21:21, 19 January 2020 (UTC) Ok sorry, I guess I should reword. I never meant the the column "booster landing" should say "went boom", what I meant was that in the description of the launch (not the landing column) we should explain what happened to the booster, in said description it is accurate to say the rocket after the abort was destroyed. In this [diff] I made said change. I guess I could have been more clear. :/ Thanks. OkayKenji (talk page) 21:40, 19 January 2020 (UTC)
 * Should we update Falcon 9 booster B1046 and List of Falcon 9 first-stage boosters to say precluded too then? OkayKenji (talk page) 21:49, 19 January 2020 (UTC)
 * I think precluded should be when unexpected event prevent landing attempt like 19 and the non existing launch between 28 and 29. This case is same as when all booster thrust is needed for mission so there is no attempt, The mission prevented it not unexpected events so no attempt. I changed it before seeing discussion here, sorry. Also slightly related matter: 2015 had 7 launches but there are only 6 shown in landings graph. Should precluded like launch 19 be treated as no attempt in graph? I agree non existing launch of AMOS-6 shouldn't have a no attempt landing for that in the graph. crandles (talk) 21:55, 19 January 2020 (UTC)
 * I agree with C-randles. Launch 19 was a failure so the landing was precluded because the rocket unexpectedly exploded, even though a landing had been planned for. The inflight abort test was a success, with the booster being expected to explode (which it did) and thus no landing was ever planned. This, to me, clearly falls in the category of there being 'no attempt' rather than being precluded (because of an unexpected event). Oska (talk) 22:58, 19 January 2020 (UTC)
 * ,, I guess the difference is that in both cases the rocket never even got to the stage of reentry, meaning that IMO, any landing was precluded. to say that no landing was attempted implies that there was a reentry, as is the case with all other 'no attempt' boosters. —  Insertcleverphrasehere (or here) (click me!)    06:32, 20 January 2020 (UTC)
 * It isn't clear why you are focusing on reentry, someone could equally infer no attempt implies there is a launch in which case launch 19 should be no attempt. So I don't see good backing for this way of categorizing which seems to me to be ambiguous and confusing. You are treating 'no attempt' as a sub category of precluded. I would say we could categorize them between 'Precluded by mission', 'Precluded by contract' and 'Precluded by unexpected events'. These 'precluded's would all be sub categories of no attempt for graph. So treating 'precluded' as a sub category of no attempt makes more sense to me than the other way around as you are doing. crandles (talk) 13:08, 20 January 2020 (UTC)
 * , Is the other 'preculded' result counted as "no attempt"? could someone do a count? —  Insertcleverphrasehere (or here) (click me!)    20:12, 20 January 2020 (UTC)
 * 2015 6 on graph total (success/failures/attempts), 7 total launches on 2015 tables
 * 2016 8 on graph total (success/failures/attempts), 9 total on 2016 tables
 * So it appears the Precluded are just not on the graph? OkayKenji (talk page) 20:27, 20 January 2020 (UTC)
 * , Yeah... for the same reason I stated above. No landing was possible, SpaceX didn't decline to attempt a landing, it just didn't happen. I thought this was the correct rationalle, but if we are going with 'no attempt', then the others should also be changed to no attempt (probably also in the list as well, and just get rid of 'precluded' altogether) —  Insertcleverphrasehere (or here) (click me!)    20:30, 20 January 2020 (UTC)

There was 7 launches in 2015, so I think all graphs should total 7 for 2015. To do this, on the landings graph we should count the precluded as a sub category of 'no attempt' and include launch 19 as a 'no attempt'. For 2016 there were 8 launches and the AMOS-6 non-launch. 9 are shown on three of the graphs but the landings graph only has 8. Given the page is a list of launches, I think all graphs should be reduced to having 8 entries for 2016. However it is possible that since rocket was built and ready to fly, some people might want to include the AMOS-6 non-launch in the rocket configuration graph only or maybe more than that one graph. Don't think I agree with that: While this means I want to include one of the precluded in the graphs (L19) and not the other precluded (AMOS-6), this is because one launched and the other didn't. We obviously don't include things that were scheduled to launch but were delayed to the following year, so I think it appropriate that it doesn't get an entry in any of the graphs unless it launched. This way the total number for each year should agree between all the graphs. Hope that makes sense. crandles (talk) 22:41, 20 January 2020 (UTC)
 * Keep Amos-6 out of the graphs, I agree. "No attempt" for the in-flight explosions (planned and unplanned) should be fine: There was no landing attempt. It doesn't say why there was no landing attempt, but for these details we have the table. --mfb (talk) 00:59, 21 January 2020 (UTC)
 * Sorry should have noted launch outcome graph has category for 'Loss before launch'. Excluding that gives too good an impression. Perhaps that should stay so that graph alone has 9 but other graphs have 8 for 2016? crandles (talk) 09:40, 21 January 2020 (UTC)
 * , I'd say so. —  Insertcleverphrasehere (or here) (click me!)    11:17, 21 January 2020 (UTC)

Seems agreement on graphs so landings increased to total of 7 for 2015, launch sites and rocket configurations reduced to 8 for 2016. crandles (talk) 11:55, 21 January 2020 (UTC)

Remaining issue seems to be whether to get rid of precluded in tables as Insertcleverphrasehere suggests or whether we keep it. There are 3 instances: L19 landing column, Amos-6 launch col and Amos-6 landing column. I think it is extra suitable information for tables and should be kept. There is perhaps an argument that Amos-6 landing column currently saying 'Precluded (drone ship)' should be expanded to indicate there was no launch (as explanation for it not being included in graph) but the adjacent launch column does say 'Precluded (failure pre-flight)' so maybe clear enough without any changes being needed. crandles (talk) 12:07, 21 January 2020 (UTC)
 * , Amos-6 wasn't a launch, so that makes sense; it simply doesn't appear on the graph (the other option would just be to write N/A in the field). As for flight 19 and the IFA, I'm still in two minds. To be the most accurate we would put something like 'preculded', 'N/A', or 'destroyed in flight' in the chart for these as well, as landing was both not attempted and not possible. I suppose we can just go with 'No attempt' as well though, as this is also technically accurate. It's not a perfect solution, but neither is needlessly sub-categorising everything and at least gives us consistency between the graphs and list. I'm still kinda in two minds about this and would bow to consensus either way. —  Insertcleverphrasehere (or here) (click me!)    20:30, 21 January 2020 (UTC)
 * I think "precluded" is trying to explain too much, in a field that doesn't have space for that. "No attempt" is perfectly true, "precluded" is trying to explain why there was no attempt. This explanation should go into the description of the launch. Tercer (talk) 08:55, 22 January 2020 (UTC)
 * Definition of preclude: "prevent from happening; make impossible". So the meaning that 'there was going to be a landing but something prevented it' seems fairly clear from the one word precluded and the description explains what did happen. What is difficult or too much? crandles (talk) 11:46, 22 January 2020 (UTC)
 * I'm not saying that it is a matter of clarity, indeed "precluded" is perfectly clear. I'm just saying that this is out of place there. In this column we should describe what happened (as is the case with all other categories), not why it happened. "No attempt" is what happened, "precluded" is why it happened. Also, it would be nice if the categories in the list matched the categories in the graph, as I suggested before. Tercer (talk) 12:39, 22 January 2020 (UTC)
 * I would say that 'what happened' can or even does include that a landing was 'planned but didn't occur'. The reason why (explosion during ascent/before launch) is to be restricted to the description. If this 'planned but didn't occur' can be fitted in with sufficient clarity, why not? There is certainly some merit in matching the categories between graph and list, and I am not particularly opposed to this esp if it is the majority view. However, with the footnotes a long way away, I prefer the controlled and uncontrolled descriptions for the ocean tests. crandles (talk) 14:42, 22 January 2020 (UTC)
 * I just checked the article, and in all ocean cases the description in the list explains well what actually happened in the test, much better than the footnotes can. So I don't think that's a problem. Tercer (talk) 15:50, 22 January 2020 (UTC)

Should the Crew Dragon IFA test Falcon 9 flight be given a sequence number in the list?
As one of those members who have been obsessed with keeping spaceflight related lists here updated for years, this Crew Dragon IFA test is unusual in that the Falcon 9 is almost complete, but happen to be missing the 2nd stage engine that would make it into a "full" Falcon 9. It's more "complete" than, say, Ares-I-X, but not quite a full rocket on a sub-orbital flight as we have seen with the first test flights of the Soyuz-2 and Angara.

So, should this flight of B1046.4 be listed as the 79th F9 flight in the list, or should we list it as "N/A" in the sequence number? The situation is so deep into gray area that I don't have any horses in this fight, so discuss away! Galactic Penguin SST (talk) 18:11, 19 January 2020 (UTC)
 * , I think we should just list it as 79. Sure it doesn't have a second engine, but that's just because it wasn't needed for this particular mission. It's still a falcon 9 (otherwise it wouldn't be on the list at all). —  Insertcleverphrasehere (or here) (click me!)    19:02, 19 January 2020 (UTC)
 * If the title was 'List of Falcon 9 orbital launches', it wouldn't deserve a number as not orbital. I think it is a Falcon 9; Missing unnecessary parts like landing legs doesn't make it something other than a falcon 9. So missing S2 engine that wasn't necessary for this mission shouldn't either. It was a launch so deserves a number when title is 'List of Falcon 9 and Falcon Heavy launches'. Loss pre launch didn't get a number, but this was a launch. crandles (talk) 17:16, 22 January 2020 (UTC)

Starlink numbering
Per, assuming this can be believed, it appears SpaceX may be numbering the launches with v0.9 as number 1. Is this sufficient to renumber them? crandles (talk) 21:33, 26 January 2020 (UTC)
 * Hmm, it appears Starlink-4 is for external use but internally they use StarlinkV1.0-L3 per [patrick.af.mil/Portals/14/Weather/L-2%20Forecast%2027%20Jan%20Launch.pdf?ver=2020-01-25-114021-193] asking for confusion? crandles (talk) 21:46, 26 January 2020 (UTC)
 * It's called 4 only publicly to lessen the confusion of less informed members of the public. Internally they call it launch 3, we call it launch 3. Our best sources call it launch 3, it's less confusing to call it launch 3. AndrewRG10 (talk) - 7:09, 27th January 2020.
 * yeah SpaceX really needs to clarify their numbering scheme, looks like its causing a lot of confusing. But since a lot of references and the Spaceflight community refers to this a "Starlink-3" it may be best to use "Starlink-3" OkayKenji (talk page) 01:07, 28 January 2020 (UTC)
 * , This really isn't anything new lol. Pretty much everything they make that has a numbering scheme has been confusing as hell. —  Insertcleverphrasehere (or here) (click me!)    01:25, 28 January 2020 (UTC)
 * Yep like the Falcon 9 Block numbering right? OkayKenji (talk page) 01:47, 28 January 2020 (UTC)
 * , version# or block#? XD —  Insertcleverphrasehere (or here) (click me!)    02:18, 28 January 2020 (UTC)