-
#860 by abaddon on 07 Feb, 2020 19:21
-
So you're saying you agree that there is zero chance - none at all - that the IFA might have shown similar issues as the OFT showed.
So, to be clear, you're good with @rpap asserting "the software failures [...] would not have been revealed by such a test" but don't like my "almost certainly would have been revealed"? This seems a little inconsistent.Come on guys. What in the world does IFA have to do with the current problem? Almost nothing. The software failures inside of Starliner exist independently of any abort test, and would not have been revealed by such a test.
This is obviously incorrect. The Starliner would, at separation, have had the +11 hour mission clock and almost certainly would immediately have behaved incorrectly following separation from the Atlas V.
Playing devil's advocate here, but you have no basis for asserting that "obviously incorrect" statement.
It's not really inconsistent. Read the rest of my reply. Software can have different state machines depending on where in flight they are. Ascent state logic can have nothing in common to nominal logic after a nominal release from the LV after a nominal insertion.
I'm going to have to disagree with you on that. -
#861 by Rocket Science on 07 Feb, 2020 19:33
-
Both NASA and Boeing should have known the spacecraft GNC alignment status during the polls for Go for launch while still on the pad... Hello Starliner, do you know where you are? Doesn't pass the smell test...
https://www.nasa.gov/feature/how-the-mission-is-controlled-inside-nasa-and-boeing-joint-operations -
#862 by mlindner on 07 Feb, 2020 19:36
-
Press conference started: https://www.nasa.gov/nasalive
-
#863 by cebri on 07 Feb, 2020 19:38
-
Bridestine "We had a lot of anomalies during OFT". Is he referring to the ones already disclosed or additional ones?
- Two issues disclosed just symptoms of the root causes.
- More fundamental problems.
-
#864 by mlindner on 07 Feb, 2020 19:41
-
Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
-
#865 by TorenAltair on 07 Feb, 2020 19:42
-
Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
same here -
#866 by abaddon on 07 Feb, 2020 19:42
-
Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
Yes, and at one point I had to refresh and start the audio over again. Weird.
[EDIT] I think the refresh thing on my end was probably a placebo and it was just a long drop out. -
#867 by cebri on 07 Feb, 2020 19:45
-
- Time Issue. Coding requirements were not met.
- Service Module Disposal. The firing sequence had been coded for free flight, should've coded (separate lookup table) for a different firing sequence. They found that issue and uploaded a fix to have a successful reentry. The risk was having the service module impacting the capsule. -
#868 by John Santos on 07 Feb, 2020 19:46
-
Early on, IIRC, it was claimed that the CST-100 initializes its MET from the Atlas at spacecraft separation (NOT at liftoff or at some point during the countdown.) Does the CST abort system need to know the MET in order to perform an abort correctly? I would think so!
So you're saying you agree that there is zero chance - none at all - that the IFA might have shown similar issues as the OFT showed.
So, to be clear, you're good with @rpap asserting "the software failures [...] would not have been revealed by such a test" but don't like my "almost certainly would have been revealed"? This seems a little inconsistent.Come on guys. What in the world does IFA have to do with the current problem? Almost nothing. The software failures inside of Starliner exist independently of any abort test, and would not have been revealed by such a test.
This is obviously incorrect. The Starliner would, at separation, have had the +11 hour mission clock and almost certainly would immediately have behaved incorrectly following separation from the Atlas V.
Playing devil's advocate here, but you have no basis for asserting that "obviously incorrect" statement.
It's not really inconsistent. Read the rest of my reply. Software can have different state machines depending on where in flight they are. Ascent state logic can have nothing in common to nominal logic after a nominal release from the LV after a nominal insertion.
I'm going to have to disagree with you on that.
If Boeing was to perform an IFA test, would it use an actual Atlas V, or like Orion (and Mercury and Apollo), would it use a much cheaper substitute rocket? And if so, would that rocket use an actual Atlas V avionics package, or would it use an emulated one? Because if so, the emulated avionics might have been written to implement what the CST-100 expected to see from the Atlas (and provided a correct MET at separation) rather then what Atlas V actually does. In the former case, the IFA test would not have revealed the MET problem. In the latter case it would. Unless the abort software is using GPS or some other mechanism to determine altitude and velocity and doesn't care what the Atlas thinks the MET is. In which case, they should DEFINITELY have run an IFA test...
Edit: during the conference, they said the spacecraft was supposed to pull the time from the Atlas during the terminal count. (It seems reasonable to assume that if there is a hold or abort during the terminal count, they will at a minimum recylce to T-4:00 which should cause the spacecraft to pull the MET again.) So my issue of the IFA not replicating the MET problem due to an abort test rocket emulating what the spacecraft expected rather than emulating what the Atlas actually does is moot. But would an abort test go through a full Atlas V countdown, which is otherwise completely unnecessary with a solid fuel rocket, or would the count down just consist of "power up Big Joe 2", "power up CST-100", "rocket is go", "spacecraft is go", "range is go", "okay to launch", "10, 9, 8, 7, ..." and the MET getting initialize a few seconds early rather than 11 hours early?
-
#869 by mlindner on 07 Feb, 2020 19:47
-
"The service module could have bumped back into the orbiter after separation." Wow...
-
#870 by mlindner on 07 Feb, 2020 19:51
-
Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
This is getting really frustrating. It appears NASA IT can't do software either. -
#871 by cebri on 07 Feb, 2020 19:52
-
Problem with MET: Two requirements: pull the information from the launcher, and pull it during terminal count. The code didn't meet the second requirement.
Disposal burn: The firing mapping was coded incorrectly. It was the same as free flight. -
#872 by cebri on 07 Feb, 2020 19:54
-
Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
This is getting really frustrating. It appears NASA IT can't do software either.
I'm actually not having any issues, have you tried with a different browser? -
#873 by abaddon on 07 Feb, 2020 19:55
-
I just switched from Chrome (had issues) to Firefox and have not had issues since.Is anyone else getting audio drop outs? I keep losing a couple seconds of audio every minute or so.
This is getting really frustrating. It appears NASA IT can't do software either.
I'm actually not having any issues, have you tried with a different browser? -
#874 by cebri on 07 Feb, 2020 20:00
-
Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
-
#875 by mlindner on 07 Feb, 2020 20:02
-
Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
A million lines is not that huge considering the project size. Your video player has about a million lines of code. The Linux bare kernel has 27.8 million lines of code, including nothing else in the OS. -
#876 by abaddon on 07 Feb, 2020 20:04
-
"lines of code" is usually a pretty terrible metric that can vary widely if they are actual code, in-line documentation, what language they're written in, etc. Hard to say what that means about the scope of the software without other info.Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
A million lines is not that huge considering the project size. Your video player has about a million lines of code. The Linux bare kernel has 27.8 million lines of code, including nothing else in the OS.
As these things go in a hand-wavy sense I agree one million lines doesn't seem like a very big number. -
#877 by mlindner on 07 Feb, 2020 20:05
-
"lines of code" is usually a pretty terrible metric that can vary widely if they are actual code, in-line documentation, what language they're written in, etc. Hard to say what that means about the scope of the software without other info.Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
A million lines is not that huge considering the project size. Your video player has about a million lines of code. The Linux bare kernel has 27.8 million lines of code, including nothing else in the OS.
As these things go in a hand-wavy sense I agree one million lines doesn't seem like a very big number.
Yes that's also a very good point. -
#878 by kevinof on 07 Feb, 2020 20:05
-
Maybe that's why it failed - they didn't put in enough error checking.
Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
A million lines is not that huge considering the project size. Your video player has about a million lines of code. The Linux bare kernel has 27.8 million lines of code, including nothing else in the OS. -
#879 by cebri on 07 Feb, 2020 20:06
-
Mulholland: All the software will be verified to check whether the requirements are being met. About a million lines of code.
A million lines is not that huge considering the project size. Your video player has about a million lines of code. The Linux bare kernel has 27.8 million lines of code, including nothing else in the OS.
And it looks like the guys developing my video player do a better job at testing. I'm baffled they didn't notice the firing sequence was coded incorrectly in such a critical moment as the separation of the spacecraft and the service module. Of all the things that I think could fail, the two issues really look like two major oversights.
