-
#700
by
deadman1204
on 27 Mar, 2023 16:43
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much 
.
Not really. Software isn't sexy so people don't think about it, but its a HUGE component of engineering. Rockets are a pile of metal without their software.
-
#701
by
Lee Jay
on 27 Mar, 2023 16:52
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
Not really. Software isn't sexy so people don't think about it, but its a HUGE component of engineering. Rockets are a pile of metal without their software.
So what? The same is true of the tanks, the avionics, the structure, the actuators, the comms, and the GSE (and everything else). It's a system that doesn't work without all the components. Equating the software with everything in a car except the tires is crazy. Software is one of many components that have to work. I'd equate the software in a rocket with the software in a car.
-
#702
by
EspenU
on 27 Mar, 2023 16:56
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
Not really. Software isn't sexy so people don't think about it, but its a HUGE component of engineering. Rockets are a pile of metal without their software.
And software is just a bunch of text without hardware. Not disputing that software is very important and often overlooked (I with with hw/sw integration). But saying it's 98% (or whatever % a car is without wheels) of a space craft is software is just delusional.
I was assuming it was a joke, but now I'm concerned that it wasn't..
-
#703
by
SoftwareDude
on 27 Mar, 2023 17:26
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased,
but software these days is the critical path even for cars.
-
#704
by
clongton
on 27 Mar, 2023 17:38
-
All the parts of an intricate model, for example, are really, really important. But without the glue it's just a kit in a box. Flight software is the glue.
-
#705
by
Lee Jay
on 27 Mar, 2023 20:22
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased, but software these days is the critical path even for cars.
So are tires.
-
#706
by
Lee Jay
on 27 Mar, 2023 20:23
-
All the parts of an intricate model, for example, are really, really important. But without the glue it's just a kit in a box. Flight software is the glue.
And without the kit, the glue is just a sticky mess waiting to happen.
-
#707
by
John Santos
on 27 Mar, 2023 20:39
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased, but software these days is the critical path even for cars.
So are tires.
"Where we are going, we don't need roads (or tires.)"
-
#708
by
SoftwareDude
on 27 Mar, 2023 21:22
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased, but software these days is the critical path even for cars.
So are tires.
"Where we are going, we don't need roads (or tires.)"
What I was seriously trying to say, is that software, for a project like a spacecraft, is at least partly, an unknown quantity. That's why we have agile development and continuous development. We don't know enough when we start writing, about what it's going to take to finish writing the code. NASA should know better; just sayin'.
-
#709
by
Jim
on 27 Mar, 2023 21:31
-
NASA should know better; just sayin'.
Not NASA doing the work
-
#710
by
Lee Jay
on 27 Mar, 2023 23:44
-
NASA should know better; just sayin'.
Not NASA doing the work
Plus, wasn't a lot of the trouble attributed to issues with testing the code?
-
#711
by
SoftwareDude
on 28 Mar, 2023 00:20
-
NASA should know better; just sayin'.
Not NASA doing the work
The context of that is based on previous remarks; I was referring to NASA's lack of oversight. They figured the Starliner's hardware was under control, so we're good. Wrong!
-
#712
by
Cherokee43v6
on 28 Mar, 2023 00:21
-
NASA should know better; just sayin'.
Not NASA doing the work
Plus, wasn't a lot of the trouble attributed to issues with testing the code?
Not sure I would term 'failing to do a full up, integrated software test of the full system' to be an 'issue with testing the code'. To me, an 'issue' would be something prevents them from performing the test. In this case, the company decided they did not
need to do the integrated test. That's not an issue, it is a project management failure.
-
#713
by
Jim
on 28 Mar, 2023 00:35
-
The context of that is based on previous remarks; I was referring to NASA's lack of oversight. They figured the Starliner's hardware was under control, so we're good. Wrong!
Wrong. This is a commercial program, there is no oversight, just insight. But either way, NASA would still not have been able to find the problems.
-
#714
by
Lee Jay
on 28 Mar, 2023 00:38
-
NASA should know better; just sayin'.
Not NASA doing the work
Plus, wasn't a lot of the trouble attributed to issues with testing the code?
Not sure I would term 'failing to do a full up, integrated software test of the full system' to be an 'issue with testing the code'. To me, an 'issue' would be something prevents them from performing the test. In this case, the company decided they did not need to do the integrated test. That's not an issue, it is a project management failure.
Yeah, I'll go along with that.
-
#715
by
FishInferno
on 28 Mar, 2023 00:44
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
Not really. Software isn't sexy so people don't think about it, but its a HUGE component of engineering. Rockets are a pile of metal without their software.
Just adding to this, in my university courses they teach that the Apollo Program/Space Race was the origin of software engineering as a discipline. Not 100% true, but spaceflight and software have always been wholly intertwined.
-
#716
by
Lee Jay
on 28 Mar, 2023 00:53
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
Not really. Software isn't sexy so people don't think about it, but its a HUGE component of engineering. Rockets are a pile of metal without their software.
Just adding to this, in my university courses they teach that the Apollo Program/Space Race was the origin of software engineering as a discipline. Not 100% true, but spaceflight and software have always been wholly intertwined.
That's true of nearly everything today (and in the last at least 4 decades). Even things without software in them were almost certainly designed or manufactured using software.
Software is part of everything, but it's just as useless as all the other components of nearly everything without all the rest of the system. Got the best rocket software in the world? Try running it without avionics or control hardware to run it on.
-
#717
by
meekGee
on 28 Mar, 2023 01:23
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased, but software these days is the critical path even for cars.
That's an indication for the state of software engineering management more than anything else.
Somehow they're always gating, and it's not because writing software is harder than developing hardware.
-
#718
by
SoftwareDude
on 28 Mar, 2023 03:19
-
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
I know software is very important, but it seems like you might be elevating it a little bit too much .
I admit I am a little biased, but software these days is the critical path even for cars.
That's an indication for the state of software engineering management more than anything else.
Somehow they're always gating, and it's not because writing software is harder than developing hardware.
It is what it is. Maybe software development could be better, but it is not, and everyone knows it.
-
#719
by
woods170
on 28 Mar, 2023 11:09
-
No offense Chuck, but SNC's lack of progress on Cargo Dream Chaser is not tied to them not getting a CCtCAP contract. <snip> They were exactly what Gerst said they were: not far enough along in development compared to SpaceX and Boeing.
I don't disagree however Boeing was obviously not as far along as they publicly purported to be and NASA (knowingly?, congressional pressure?) bought the lie. One need only look at the very sad state of affairs wrt their competence the way Starliner performed on both its test flights. And the 2nd one was not without problems either, that they are STILL working through, as demonstrated by the crewed flight being delayed - yet again - to fix -another- problem. And they are STILL redesigning that sticking valve! IMO, the true state of "far enough along" had Dream Chaser and Starliner approximately equal, at different places to be sure, but overall approximately equal. In addition, there was a LOT of pressure to make sure Boeing got a contract. Without that silent requirement, SNC "might" have gotten the nod. That is the source of my "cheated" statement.
You are overlooking a few things. From all perspectives, bar one (software), Boeing was ready to fly the uncrewed test flight just 9 months after SpaceX. And that was for a company that DIDN'T have the benefit of building off a prior generation of capsules.
And NASA, by voice of Kathy Lueders and others, have already admitted that they didn't pay as much attention to Boeing's software efforts as they should have. They were pre-occupied with SpaceX, because SpaceX used a software development standard with which NASA had no prior experience. That led to NASA being so busy trying to understand what SpaceX was doing, that they "trusted Boeing a bit too much", based on prior experience.
But in 2014, when CCtCAP was awarded, it was clear to NASA that Boeing and SpaceX were both well ahead of SNC. The call Gerst made back then was the right one.
WIthout the software, its like saying they are building a car and have the tires figured out; all that is left is the car.
That is overstating it by a fairly wide margin. But I get the idea.
As Kathy Lueders, and others, admitted back in 2021, their focus WAS on software, just not Boeing's. They also admitted that THAT was a mistake. One that they have now fully corrected. Which is why Boeing has still not gotten around to flying CFT.
but software these days is the critical path even for cars.