SpaceX Falcon 9/Dragon CRS SpX-1 MISSION GENERAL DISCUSSION

[majormajor42]

So, this sentence in the article

During that first approach in May, the Dragon was working from a model of the ISS that wasn’t totally accurate, as pieces have been added to and subtracted from the real-life station.

...was it said by Elon? It is not in quotes. Are they Elon's words or Joe Pappalardo's, the author?

[Idiomatic]

I've been in software development my entire career and new software generally doesn't mean modified, upgraded software with tested fixes. It generally means rewritten software. So, to me, the original statement that no new software is required is a reasonable statement. Somebody is itching for an argument where no ground for such exists.

Sorry, but every new line of code IS NEW SW. Yes, that new Sw module gets tested again, just like the old code that was replaced. There may be no new functionality, but you still need regression testing.

You aren't a programmer. No coder I know would call a patch over a hotfix 'new software'. Elon Musk is a programmer from the start. He likely would be of the same mindset as other programmers on this issue too.

[A_M_Swallow]

I've been in software development my entire career and new software generally doesn't mean modified, upgraded software with tested fixes. It generally means rewritten software. So, to me, the original statement that no new software is required is a reasonable statement. Somebody is itching for an argument where no ground for such exists.

Sorry, but every new line of code IS NEW SW. Yes, that new Sw module gets tested again, just like the old code that was replaced. There may be no new functionality, but you still need regression testing.

You aren't a programmer. No coder I know would call a patch over a hotfix 'new software'. Elon Musk is a programmer from the start. He likely would be of the same mindset as other programmers on this issue too.

Regression testing performed at 3 a.m. on a Sunday morning when the modified program will be doing a production run at 3:15 a.m. is very limited.

[beancounter]

NASA will want to review all the software changes, every flight, thus the strong desire to claim that there hadn't been any.

Really?!!  NASA has qualified SpaceX to commence their CRS contract.  Why would they now require oversight on a commercial company's software? 
Can't see it happening.  They would have had to have this written into the CRS and since it's a cargo supply contract, it's reasonable to expect that it would only deal with that.
Software would be part of a development contract i.e. COTS, which is now complete.

I didn't say anything controversial.. NASA reviews the software of all visiting vehicles.

Do they?  Well in development, test, etc, but in under normal operations?  Source please?  I presume this also applies to HTV ATV Soyuz?  Haven't seen this mentioned anywhere before. 

[QuantumG]

Do they?  Well in development, test, etc, but in under normal operations?  Source please?  I presume this also applies to HTV ATV Soyuz?  Haven't seen this mentioned anywhere before. 

Yes. I heard it at the press conference during COTS2+, but I'm sure someone can provide you with a link to the visiting vehicle requirements.

[beancounter]

Do they?  Well in development, test, etc, but in under normal operations?  Source please?  I presume this also applies to HTV ATV Soyuz?  Haven't seen this mentioned anywhere before. 

Yes. I heard it at the press conference during COTS2+, but I'm sure someone can provide you with a link to the visiting vehicle requirements.

Yep appreciate that.  Be very surprised if it includes NASA oversight on any existing vehicle with agreements (contracts) in place.  Mind you, guess the term 'oversight' could be construed in both general or very specific terms.  I'd interpret it in very general terms, ie. tests must have been done, etc.

[QuantumG]

Yep appreciate that.  Be very surprised if it includes NASA oversight on any existing vehicle with agreements (contracts) in place.  Mind you, guess the term 'oversight' could be construed in both general or very specific terms.  I'd interpret it in very general terms, ie. tests must have been done, etc.

It seems you're very surprised over everything.

Yes, NASA reviews the code in the Soyuz, and the Progress, and the ATV and the HTV. It's a requirement of bringing your vehicle into the exclusion zone of the ISS. Yes, that includes every software change, otherwise what would be the point?

This is not news.

[ChefPat]

How much software can dance on the head of a pin?

[Lurker Steve]

I've been in software development my entire career and new software generally doesn't mean modified, upgraded software with tested fixes. It generally means rewritten software. So, to me, the original statement that no new software is required is a reasonable statement. Somebody is itching for an argument where no ground for such exists.

Sorry, but every new line of code IS NEW SW. Yes, that new Sw module gets tested again, just like the old code that was replaced. There may be no new functionality, but you still need regression testing.

You aren't a programmer. No coder I know would call a patch over a hotfix 'new software'. Elon Musk is a programmer from the start. He likely would be of the same mindset as other programmers on this issue too.

I'm not ?? Does that mean I can go home now ?? I'm so confused about what I've been doing the last 35 years of my life. (Yes, I've been doing software development MUCH LONGER than Lord Elon Musk).

There may be different review and testing processes for source code changes less than 5 or 6 lines, versus a replacement of an entire module. Either way, reviews and testing are going to be required if your organization has any type of structured development process in place.

[Idiomatic]

I never said they wouldn't review and test the patch. I only was speaking of pedantics. A complete module rewrite could probably be called new software, an actual new module would almost certainly be called new software, a patch, even hundreds or thousands of lines ... probably not. Anyways, I'm done, I wasn't debating you so much as Quantum's fight picking.

[friendly3]

I'm so confused about what I've been doing the last 35 years of my life. (Yes, I've been doing software development MUCH LONGER than Lord Elon Musk).

Lord Elon Musk Vader has been doing rocket development MUCH LONGER than you !

[beancounter]

Yep appreciate that.  Be very surprised if it includes NASA oversight on any existing vehicle with agreements (contracts) in place.  Mind you, guess the term 'oversight' could be construed in both general or very specific terms.  I'd interpret it in very general terms, ie. tests must have been done, etc.

It seems you're very surprised over everything.

Yes, NASA reviews the code in the Soyuz, and the Progress, and the ATV and the HTV. It's a requirement of bringing your vehicle into the exclusion zone of the ISS. Yes, that includes every software change, otherwise what would be the point?

This is not news.

Ok so every vehicle?  A full review of all onboard software code?

[QuantumG]

Ok so every vehicle?  A full review of all onboard software code?

Anything that changed.. that's why they have strict change control.

[beancounter]

Ok so every vehicle?  A full review of all onboard software code?

Anything that changed.. that's why they have strict change control.

I understand change control.  So what we've got to is that NASA has not only oversight but the right to fully review any software changes that are made.  No changes, no review.  Btw, I'd consider right to fully review somewhat further along than simply oversight but it's a small point.
And no need for the narky, some may interpret as rude response.

[QuantumG]

I understand change control.  So what we've got to is that NASA has not only oversight but the right to fully review any software changes that are made.  No changes, no review.  Btw, I'd consider right to fully review somewhat further along than simply oversight but it's a small point.

And no need for the narky, some may interpret as rude response.

Sorry, I'm baffled as to why we're still having this discussion.

This has nothing to do with "oversight". It's a requirement of the ISS visiting vehicles standards. No-one is allowed into the exclusion zone without meeting those standards.

[beancounter]

I understand change control.  So what we've got to is that NASA has not only oversight but the right to fully review any software changes that are made.  No changes, no review.  Btw, I'd consider right to fully review somewhat further along than simply oversight but it's a small point.

And no need for the narky, some may interpret as rude response.

Sorry, I'm baffled as to why we're still having this discussion.

This has nothing to do with "oversight". It's a requirement of the ISS visiting vehicles standards. No-one is allowed into the exclusion zone without meeting those standards.

Well think it started out as NASA 'oversight' but has now turned into 'standards'.  Oh well.  As you say, why bother. 
Looking forward to the first CRS flight but lost my bet on 3 flights this year.  Still 2's better than none.

[QuantumG]

Well think it started out as NASA 'oversight' but has now turned into 'standards'. 

Huh? The visiting vehicle standards were written before the commercial cargo program was even imagined. Who do you imagine they were overseeing?

[oldAtlas_Eguy]

Well think it started out as NASA 'oversight' but has now turned into 'standards'. 

Huh? The visiting vehicle standards were written before the commercial cargo program was even imagined. Who do you imagine they were overseeing?

Since the F9 software for ascent is run on the computers of the F9 US and the Dragon is not a active participant, any changes of the F9 software would not be reviewed by NASA due to the VV "standards".

Only the spacecraft software must be shown as not having a safety issue. Not all tests would be done for a given software change on the Dragon either. Most software changes would only involve a small subset of the complete test set.

BTW this is from my experience being the Chairman of the AF Software Review Board for the Atlas E/F guidance software.

[QuantumG]

Since the F9 software for ascent is run on the computers of the F9 US and the Dragon is not a active participant, any changes of the F9 software would not be reviewed by NASA due to the VV "standards".

We were talking about the software on the Dragon.

[simonbp]

How much software can dance on the head of a pin?

Well, say a typical NAND flash chip is about 1 cm2 and has a capacity of 32 GB. The head of a pin is about 1 mm2, so that's about 32 MB. Allowing for file system overhead and whatnot, that's about 30 MB of software on the head of a pin.

Out of curiosity QuantumG, have you ever actually developed software yourself?