That's a no op for me since I've always avoided using switch/case statements anyway. I always saw switch statements as an omission of poor design...

Please describe a construct that you can do in C that you can't do in python?

He could have simply made the book 260 pages instead of 240 and put in a 20 page chapter on ZFS right after RAID. The first couple of pages would be about the design philosophy of ZFS. Next introduce the concepts of vdevs, pools and pool types (in relation to what the reader just learned about RAID), sub file systems, snapshots and file system attributes. Next layout some scenarios using 8 disks in a JBOD. Create a raidZ, raidZ2 and a raid10. Next talk about tacking on another 8 disks and what the options would be for expanding a raidZ, raidZ2, raid10 set. Next talk about the pros and cons of read caches and ZIL's and ways to tune ZFS to be more performant. Lastly, talk about scrubbing and replacing failed devices.

I'll stand by my original argument... ZFS is essential to building scaleable networked storage devices with FreeBSD/Solaris and likely soon Linux. Yes, you could write the end all book on ZFS. Yes, someone like me would likely buy such a book. However, for your average sysadmin who knows nothing about ZFS this chapter plus google would give them a good starting foundation for building a storage device.

Naming a book "Storage Essentials" and then not talking about ZFS was a mistake. If you're going to be building any type of NAS, you're going to want to use ZFS for it's scalability, reliability and stability. While you might get away with UFS for a couple of terabytes, you're going to have a bad time of it when you've got 40TB worth of storage space to manage.

SpaceX has modeled the hell out of it. It's just really really hard. Honestly, I'd rather they fail in spectacular fashion and explore all the dark corners of their design before they stick people on the top of it. What's great about this is it's all gravy at this point. Once they work all the kinks out, it's going to eviscerate the competition when it comes to cost to orbit per ton. I can only imagine every other commercial launch company must look at SpaceX with a mix of horror and amazement.

Yeah, because in the 60's we were building reusable rockets that were autonomously landing on robotic ships in the ocean... oh wait.

Every magazine and newspaper in France should publish the offending cartoons on their front pages. Show these assholes that they wouldn't silenced with violence.

Agreed. I think the couple of days is due to orbital mechanics more than anything else. It could also be that because this resupply is so critical they're going to do some additional testing.

From what I know about SpaceX their testing regime is pretty insane already.

I've read both the offending article and the response from Krauss and frankly Krauss is right on the money. The article is so painfully full of woo and so devoid of fact I can only come to the conclusion that the editors at the WSJ are a bunch of biased religious pandering idiots. What's even more enjoyable is how the refused to print his rebuttal because in doing so it would have show how painfully shotty their editorial process is.

Dr. Krauss has done us a service by clearly demonstrating the WSJ is good for nothing more than lining the bottoms of bird cages where it can get treated with the respect it fully deserves.

Recreating my machine from install media isn't that gruesome either. However, I'd rather do it on my terms then have to suddenly deal with it. Murphy's law dictates it'll happen two days before a deadline or in the middle of something critical.

All the stuff I care about I make incremental offsite backups.

If you find yourself resorting to baking your motherboard in the oven to fix it, this is an act of desperation. Yeah the guy managed to get some more life out of the motherboard but this isn't a fix, this is a temporary hack.

We've now conflated two important distinctions into a single subject here. Functional resilience and long term data integrity.
I solve the long term data integrity problem by doing nightly snapshot delta's of my whole machine and my wife's machine (to a rasp pi with an external drive at a buddies house). Granted that's a single point of failure, but it's out of house in case my house {burns down, get's robbed, etc}

However, that doesn't fix the near term issue of me busily working away on a project when boom, my drive fails and suddenly I'm sitting there looking at a paper weight. That sucks. Having that happen to me once was enough for me to say screw it, I'm buying two drives and mirroring them using the motherboard raid software (which md supports) and it's a non problem. This solves my functional resilience.

raid is not a backup, what is gives me is resilience. Would you rather spend tomorrow recreating your machine from install media and backups, or simply swapping the drive out and suffering a background sync?

I don't build a machine these days that doesn't have mirrored hard drives. You realistically can't backup 6TB worth of data, so barring some horrible FS failure (which is rare these days in Linux land) your best bet is RAID1.

For a falcon heavy having a fleet of these ships would allow them to recover all three first stage elements vastly reducing the cost to put really heavy things like all the elements for a mars transfer vehicle in space.

I agree plus it gives them flexibility. As much as Elon talks about having the first stage land back at the launch site, there's simply too much valuable infrastructure. Having a fleet of these autonomous ships gets Elon the reusability without endangering his personal and launch sites.