Hacker News Reader

Comment by _ph_ | Hacker News Reader

From what I picked up, their fuel budget for the landing was borderline, as this satellite mission didn't leave them much. This means, they had to go in very fast, and also use a 3 engine landing burn instead of the usual 1 engine. So it was more a test whether they still can land the rocket rather than a proper landing. For that, it performed well.

View on HN

Yup. And Elon had queued up that exact talking point roughly 3 hours before the launch. However, having watched this company for a while and Elon in particular, he is much more nuanced in what he says. All through SpaceX's life they have leveraged the customers expectations in order to do their own R&D on the customer's nickel. So when they weren't bringing back rockets they charged for the cost of the rocket and then on the now 'waste' rocket invested in using it to test recovery systems of various kinds. Losing the test vehicles is not a problem its cost was covered by the launch. Now that they have some pretty good understanding of recovery, they use 'flight proven' boosters as their test vehicles, bringing them in under different flight regimes using different recovery parameters. I really think it is a brilliant strategy and one that I admire. What I think about is if I was working for them the even more stuff I'd be doing to get even more data from the tests. (like a better video recovery system, which truth be told they very well may have, after all they are under no obligation to show us and their potential competitors what they are trying to improve and what they don't know)

> (like a better video recovery system [...])

I can't tell by your wording, and I was similarly confused about their video recovery system, so maybe this will be helpful for you.

The only thing SpaceX hasn't invested in is a better downlink. The video is fully recorded and stored on OCISLY, so they will analyze and post later.

The rocket merely interrupts the realtime feed.

Which leads me to believe SpaceX just doesn't care about a robust realtime feed since they can analyze offline -- or they just don't care about one for the public.

If that video exists I'd love to see it some day.

Given that the downlink is reasonably stable when OCISLY is just sitting there (with our without a booster sitting on it) I'd love to see them back up on the locally recorded video and then send it via the downlink once the link has restabilized. We're talking about 18 seconds of video here, not hours :-)

I would be interested to hear how SpaceX internally views sharing video with the public. I could imagine a range of feelings from "why waste time showing them" to "an essential part of our marketing and outreach." I personally see it as fascinating stuff that touches on so many interesting questions about rocket design, future space feasibility, and the design limits around cost optimization.

Usually SpaceX does share the video after the fact, so there's some precedent. Although on occasion they have opted not to share the video as well.

So time will tell. :)

I mean, it's not like a really robust realtime feed would be much help for them anyway. As far as I know the rocket is completely autonomous, so even if something was going wrong during the landing a higher-quality feed would just give them higher-quality reasons to tear their hair out etc.

I disagree with this. Having diagnosed a number of problems in nominally 'autonomous' robots through the use of video. There are also instances of many issues being diagnosed by video, consider the loss of protection tiles on Columbia which was diagnosed (or at least confirmed) by watching video of the lift-off.

That said, SpaceX has a phenomenal number of video feeds and I really enjoy them. Early in the Falcon 9 project we would see on the web cast a cut to a video shot with no explanation of what it was, but we know now that they were looking a fuel settling and other effects during re-entry.

And yes, given that you can learn a lot from said videos about how it is done, and there is at least one high profile (Blue Origin) and a no doubt a number of low profile efforts on going to duplicate what SpaceX can do, there is some competitive advantage from not sharing all the video they have.

I still want to see them though.

I'm not saying that video wouldn't be helpful, just that realtime video wouldn't be helpful. They surely record the video for later analysis as others have mentioned.

Fair point, I was just trying to say that even near real time (record then replay once the barge had stabilized) would be useful for me.

None of the benefits you mentioned requires a real-time feed. They just need to capture and store the video for later analysis.

Unless they want to, you know, analyse the landing before they launch another rocket two days later :-) . But of course that doesn't require realtime, just the ability to download via satellite while drone ship is on the move.

> The video is fully recorded and stored on OCISLY, so they will analyze and post later.

Can you link to any posted-later videos of landings? They pretty consistently either drop frames or cut out entirely on the live stream around the moment of touchdown, so it'd be great to see decent quality footage of some landings.

Reply on news.ycombinator.com