"Duress": I do not think that word means what you think it means.

Black's Law (quoted here: http://en.wikipedia.org/wiki/Duress) defines duress as: "any unlawful threat or coercion used... to induce another to act [or not act] in a manner [they] otherwise would not [or would]".

In other words, I can induce you to sign a contract with any LEGAL threat that I so please, and the contract is still binding. But if my threats are inherently illegal (such as threatening to hurt you, hurt your family, destroy your property, blackmail you), then the concept of duress applies, and you have a defense against my breach of contract claims.

This definition makes a lot of practical sense, if you think about it. If duress were a broader concept that included me refusing to provide you with services if you don't sign a contract, then I wouldn't even legally be able to tell you the point of signing the contract, in the first place. Under that kind of twisted logic, if you asked "Why should I sign this contract agreeing to pay you $20 to mow my lawn", and I responded "Because I won't mow your lawn for free", then I'd be subjecting you to duress. Clearly, that's not conducive to basic business arrangements.

In this case, the firefighters would be threatening to withold a service (fighting the fire consuming the man's house), which doesn't seem to be an illegal threat, to me. Granted, the house represents a very serious economic and emotional loss to this man and his family. I don't want to belittle that. But it's not like the firefighters set the man's house on fire, in the first place.

Now, there are some situations where a society will legally or socially obligate an individual member to act on behalf of his fellow man in a time of need. Some jurisdictions even have laws requiring you to aid another human being in distress, as long as you're not putting yourself in harm's way (like in the Seinfeld finale). So everything I said, above, assumes that this little Tennessee burg isn't one of situations.

A DVD is a tangible good, no different than a book.

Bullshit, on face. While it's true that DVDs and paper books ARE both tangible objects, there are infinitely many OTHER differences between DVDs and books. For example, here's a list:

- Books are usually mostly composed of plant-derived cellulose, while DVDs are mostly made of petroleum-derived polycarbonate plastics.
- Books were first manufactured and published in ancient times, while DVDs weren't invented until the mid-1990s.
- Books are usually published by book publishing companies, which (as an industry) have a distinct history, corporate culture, and business model in comparison to film and television studios, which usually publish DVDs.

Now, you can call some (or all) of my distinctions silly. But since you failed to provide any supporting rationale as to why the property of "tangibility of medium" is any more relevant, I think my reasons are just as good as yours. As I see it, you have basically two choices, here:

(A) Add some argument to support the relevance of tangibility, OR
(B) Re-phrase your statement by (heavily) qualifying your assertion that the two types of items are "no different".

Otherwise, my bullshit call stands.

If planned restarts of individual servers are *not* "LOW COST" operations, you should probably concentrate on fixing your architecture and procedures, rather than delving into kernel mangling.

Please, go back and re-read my original post: I make no assumptions about whether reboots are low-cost or high-cost, in the absolute sense. So, your opinion on the absolute costs of server reboots, while interesting, is entirely irrelevant to my argument.

See, I'm discussing the RELATIVE cost of server restarts. But because you started composing your reply before you read enough of my post to understand my point, you're just lecturing to yourself, here.

Why ? How is starting from an inherently less known and more volatile state ever going to be "MORE stable" ?

Again, you're just spazzing without bothering to read the entire post. To the extent that it's possible (given that we're speculating about the future, here), I already directly answered your question.

Also, note that my argument isn't premised on in-place kernel upgrades becoming MORE stable than the reboot-facilitated kind, just ALMOST AS stable (to the point where the additional risk is trivial and outweighed by the (minor) inconvenience of a reboot). So you're basically just engaging in a straw-man fallacy, here.

All systems develop runtime cruft over time and benefit from a reboot. To say nothing of how useful planned reboots are for confirming that your systems will recover from the _unplanned_ ones.

I'm sure these are interesting opinions in the context of your own personal experiences, but they're much less relevant in an abstract discussion like this. That's because my personal experiences may differ sharply from yours, just as the experiences of any particular 3rd-party observer may differ from both of ours. (In other words, why should we Big City folk care how some country boy sodbuster does things in his podunk backwater?)

So I'll refute your argument (and give you a handy little demonstration of why argument-by-anecdote is made of fail) by simply saying: In my own personal experience, it is NOT TRUE that all systems develop "runtime cruft" and benefit from reboots, OR that planned reboots are useful. (Probably because my architecture and systems management strategy differs from yours.)

So, here's a summary for your reference:

  * You need to improve your reading comprehension skills.
  * You need to control your irrelevant outbursts and focus on the specific arguments at hand.
  * You need to learn how to identify key arguments and differentiate them from side issues.

You go ahead and keep putting all that effort into denying my suggestion. If would have been far simpler for you to just drop the single line I quoted from you above...

That's not an argument, that's just the punchline.

At first glance, you appear to be the exact type of person the Bruce Schneier was trying to warn when he said:

I tell people that if it's in the news, don't worry about it. The very definition of "news" is "something that hardly ever happens." (http://www.schneier.com/essay-304.html)

Here's a little education, if you please:

Second time suggests that...

Two data points do not a trend make. People who try to force conclusions from such limited data often hold opinions of limited value.

Also, there are hundreds of reported accidents in the U.S. oil industry, every year. You (and whoever modded you "insightful") are ignorant because most accidents don't make front-page news. The vast majority spill little or no oil, and cause little or no environmental damage or economic loss, and the operating companies pay cleanup costs plus fines for what damages do occur.

This rig explosion, unlike the Deepwater Horizon incident in April, is a very MINOR oil spill, with no worker casualties and minimal economic impact. If the rig featured in this news story had exploded back in March, before Deepwater Horizon's big spill seeded us with fear of globally-catastrophic oil spills, the article would never have made national headlines, because nobody would have given two shits.

And in another few months, maybe a year, you and the rest of this over-excitable country will completely forget that they were ever scared of catastrophic oil spills.

I'm not saying that your stuff is directly at risk. But if we have another explosion, pipeline leak, or similar event anywhere within USA jurisdiction, your property values will get tarred by a very broad brush. Anyone at risk of this needs to get politicking for some kind of review that will assure potential buyers that they won't be shafted by their petrochemical neighbors.

Where did you get this idea, that the US population is on the verge of living in fear of filling stations and refineries? IF a series of massive catastrophies struck, and IF they were all confined to the oil industry, and IF it all happened near populated areas, and IF they all happened in a short enough period of time, then you might start to see property values changing. But absent that kind of chain of unlikeIy events, I don't see it.

See, the thing you might be missing is, oil spills have been happening occasionally but regularly for about the last century or so. Refinery explosions and filling stations fires, explosions, etc. are nothing new, either. And it's not like people aren't aware of them--if there's a body count, or a big economic impact, there's usually at least a local news stories. Absent some kind of new, ongoing threat that happens close to where people live, why would anyone start caring much more than they do, right now?

BTW, there is absolutely no need to lay this kind of thing off to enemy action. Not when 8+ years of ineffective oversight coupled with corporate "long term" planning that fails to look beyond next quarter's profit and loss statement are more than adequate to account for these incidents. (I was about to say "accidents", but it appears that these are far from accidental. They look much more like the productive of short term greed multiplied by long term stupidity.)

Ah. I see. You're a paranoid conspiracy nut. Sorry, go ahead, you were saying?

Get a little imagination, will ya? Here,I'll boil your objection down into two simple premises, for you:

1) In-place kernel upgrades are inherently RISKY to stability, compared with normal reboot upgrades.
2) Reboot upgrades are a LOW COST operation.

You seem to assume that the risk of #1 (upgrade in-place) will always outweigh the cost of #2 (rebooting to upgrade). At the moment, you MAY be correct in that assumption, but we have no basis for any conclusions, yet.

But Ksplice's current business plan is to get ahold of a massive, low-cost testing infrastructure by getting installed by default on as many popular Linux distros (Ubuntu, Fedora, etc.) as possible. Properly executed, a massive testing and development effort should improve KSplice's quality (read: stability) over time.

At some point, if KSplice does it right, in-place kernel upgrades will become stable enough to no longer entail measurably more risk than traditional reboot upgrades. If/When that happens, you'd be a fool to continue reboot upgrading, right? If there's no practical added risk, why should you have to even put up with the inconvenience of a single minute's delay, or the hassle of closing and re-opening all your SSH sessions?

Hell, it's reasonable to imagine that in-place upgrades could even become MORE stable than reboot upgrades (eventually). If that happens, you'd have to be more than a fool to continue rebooting--you'd have to be some kind of technical cargo-cultist, unwilling to offend the Machine Gods by departing from the correct rituals. (There will probably be at least a few of these people--I know some of them, I think.)

For another perspective, consider these:

  - guns vs. bows
  - automobile vs. horse-and-buggy
  - pen/paper vs. typewriter
  - typewriter vs computer
  - multiprocessing vs. single CPU

Reasonable people expect that the earliest incarnation of a new technology will be buggy, unstable, dirty, explosive, unreliable, or otherwise potentially hazardous. But given time to iron out the bugs, there's eventually a tipping point where the original technology no longer fulfills its basic purpose as well as the new-fangled competitor.

God, what a turd! You premise your whole fucking stupid post on the argument that because your tape backups are offline and offsite, they're superior to my mirrored RAID drives.

Too bad you didn't bother to read my post, where I mentioned that I TAKE THE MIRRORED RAID DRIVES OFFLINE AFTER MAKING A BACKUP SET and store one of the mirrored copies offsite.

Untreated ADHD fucks like you are the reason the Internet sucks. Go make some Youtube comments, you'll fit in over there.

You're not sure because you have no firsthand experience with it. You don't even have any engineering papers with original research to cite, or even a single blog post with somebody else's anecdotes.

In short, you've got jack shit, and you're just parroting the same crap received wisdom as everybody else. Thank you for proving my point, for me.

If you're concerned about the drives breaking down over time, you can periodically test the copies. My data is all checksummed by BTRFS, but you can do it manually with any number of common tools on any filesystem.

See what I mean? There's a perfectly simple, perfectly obvious solution to your complaints, but you refuse to even try to think about how it could work. You're willfully stupid, which is pretty sad for someone who claims to have a science background.

Oh, and if you're dumb enough to think that tape media is particularly long-lived, I'd suggest you read a few of the real-world studies about stored tape lifetimes. Unless you carefully control the storage environment's temperature, humidity, etc., the failure rates will eat you alive within a few years.

As for your supposed failure probabilities of stored drives: I've stored somewhere around 40-50 PATA and SATA drives in boxes under my bed, in my closet, etc. for years--in some cases, as long as 7 years. In that whole time, I've never had a single stored drive fail to spin up, or even fail within the first 24 hours of use.

Finally we'll know the answer to the question of the decade: "And then who was phone?!?"

Of course, hard disks by themselves may seem cheaper, but they are not a true archival medium. There are so many moving parts in a HDD and each of them (bearings, heads, spindles, motors, controller card) are a point of failure.

First of all, when a hard drive is turned off and unplugged, the moving parts aren't moving. So the mechanical wear and stress of constant operation aren't really a problem. If you dump a backups set to an HDD, seal it up like from the factory, and stick it back in a box on the shelf, the differences between the construction of a tape cartridge and a hard drive are irrelevant.

And even if stored hard drives were less reliable: Ever heard of RAID mirroring? Or how about just making a second copy? I bought an external tray-less eSATA hot-swap dock (http://www.newegg.com/Product/Product.aspx?Item=N82E16817153112&cm_re=esata_hot_swap_dock-_-17-153-112-_-Product) for about $55. Combined with a hotplug script and BTRFS mirroring, it's just easy-peasy. Sure, my 1 TB backup set costs me $160 (2x 1 TB SATA drives at $80)--but I didn't have to spend a couple of grand on an over-priced tape drive.

Every time I hear another IT guy parrot this line about moving parts, a little part of me dies. It's a nonsense assertion, pretty thinly supported. It's real function is to mentally justify, ex post facto, a lack of mental effort to consider better ways.

The only missing part is that Nokia really hates Perl, loves Python, or both.

Yeah, some of us are throwing a party to celebrate, next week. We were going to invite you, but... Well, no offense, but nobody like a Perl programmer.

Bit the bullet put your idealism aside and install Windows and Office and get your work done.

No, I have a better idea. You should put your ignorance away, instead. Educate yourself:

  * http://vdxtosvg.sourceforge.net/

If I get a Visio doc, I just politely ask the sender to export the VDX, instead, so I can turn it into an SVG and play with it. If I don't care about modifying the diagram, I'll just ask for a PDF.

Nobody's turned me down, yet.

And, EVEN if I did run into a situation where I couldn't ask, or the sender couldn't/wouldn't give me a usable format, why in the hell should I bother installing Windows? Sign. Again, you can educate yourself:

  * http://www.codeweavers.com/

Seriously, this is pretty basic research, you should be doing this stuff, yourself. I'm not going to help you out, next time.

Everying f***ing time I hear somebody say "But I HAVE to keep Windows, for Visio!", I thank my lucky stars that I never learned that damn thing. OpenOffice Draw isn't quite as slick, but for 99% of the shit people don't think OODraw can do, the reality is that they're just to willfully ignorant to learn how OODraw can do it. And, bonus, I don't have to deal with the cognitive dissonance of justifying keeping a $200 OS for the sole purpose of running one app of dubious uniqueness.

I dunno, I think this whole "internet" thing is kind of overhyped, at this point. It's clearly peaked in the marketplace, and public opinion is already starting to turn against it. This time next year, your precious little Internet darling is gonna be so irrelevant that it'll make Vanilla Ice look like Joel Spolsky.