Gee, it sure is a good thing the end of WWII didn't precipitate five decades of near-constant petty warfare and general unrest! Whew! Dodged that bullet!
The more you press me, the more I'm certain you've never done any kind of SA work.
See, when you plan downtime, you plan more than you need. If you think a conversion will take six hours, you plan for twelve hours of downtime. The fact that the bank migrations always finish early means that nothing -- or very little -- goes wrong. That means that bank conversions -- which are far more common than you seem to believe -- go very smoothly and are much less fraught with danger than you'd make them out to be. (You'll note that the potential for data loss or corruption during the planning stages is numerically indistinguishable from zero.) I wouldn't call them "routine," but, with ample planning (as with any migration or conversion), as routine as such a thing can get. You presented bank migrations as something between brain surgery and bomb defusement on the delicacy scale; my point was simply that, given the rate of problems I've seen second-hand, it's nothing of the sort.
So I reiterate my initial point: given the degree of delicacy and potential data loss I see in the average bank migration, my bank can change OSes fifty-two times a year if they want. I don't care.
They do a conversion roughly every three weeks. The sensitive work -- the part where there's actual downtime -- takes place in about half a day. The rest of the time is planning, doing about a zillion test conversions, and converting over what they can without downtime. With ample planning, even a conversion of this complexity is reduced to a few weeks of leisurely work and a morning of downtime. It's not like they go in on Monday morning to get the specs and walk out after lunch with the thing done; but it's also not so error-prone that I would even consider forbidding my bank from switching software or even OSes. The fact that conversions usually finish early during planned downtime shows the degree to which it is not an error-prone process, not the degree to which it is a simple process.
That's what my roommate does. He works for a company that writes banking software, doing the conversions for banks moving from other systems, frequently running Solaris or AIX, sometimes running Linux, to theirs, which runs on Windows, which is a hell of a lot more difficult than switching between Unices.
Maybe you thought I didn't read the GP post, but I did. I didn't mention a damn thing about DNS or web servers; I was talking about banking software. Pull your head out and realize that knocking down a straw man doesn't make you big man on campus.
That's a troll? Man,/. is getting weird. I miss the days when GNAA and goatse links were trolls. Seems like this guy (who has been modded up to +4 as of this post) was dead wrong on which way the./ community leans when it comes to Sun.
Dude, if you can't even figure out how to buy a service contract with your hardware, you sure as hell shouldn't be managing any system more powerful than a toaster.
Sun has started down a long road of realization that their proprietary chips aren't worth the silicon they're printed on in a marketplace that values interoperability over nearly all else. HP realized this a while back, and have all but phased out their Alpha and PA-RISC lines.
That said, there is a place for non-x86 chips. HP has replaced most of their Alpha and PA-RISC lines with x86 chips, but some of the high-end boxes went to Itanium. IBM is still pushing POWER -- hard. I don't suspect that SPARC will die, but I do strongly suspect that, in 5 or 10 years, Solaris on x86 will be the standard and Solaris on SPARC will be the rare beast, still grinding away in a handful of data centers on Wall Street, right next to an HP box running Non-Stop OS.
However, depending on where Sun takes Solaris on x86, they could turn it into a huge market. They're already working on certifying most HP server hardware and, AFAIK, have their sights set on the server hardware of other major manufacturers as well. If Sun could make Solaris as easy to install anywhere as Linux or NetBSD (okay, so that's a bit of stretch), and if they keep giving it away for free, they could see sizable growth in their OS market.
The problem, of course, is that it's hard to make money off of a free OS, but, if you've been listening to Sun's web expos over the past year or so, you know that they label Redhat, somewhat strangely, as one of their primary competitors. Sun is obviously trying to stress their free OS; my question is whether or not they're trying to go too many ways at once. Sun provides not only a free OS, but also tons of enterprise software for it, plus hardware based on two different processing platforms. They're competing with Redhat, Novell, and HP all at once, and that's tough to do without stretching one's self too thin.
My prediction: if Sun is still a player in 10 years, it will be a very different Sun than we see today.
Oy! I understand that/.ers might not always RTFA, but can't you at least read the goddamn summary? It's a friggin' paragraph, it's not like you need to be in Mensa to understand it.
In the sense that the article is essentially an overlong rant no better or worse than the usual slashdot missive, and that it's on the home page right now, it did at least get past one "editor." Let's see... Ahhh yes, that would be Taco.
Don't worry; Zonk will dupe it before the day's over.
It's absolutely amazing how quickly all of the posts pointing out that this story is a dupe get modded down. Mindblowing that people would be wasting their mod points on a story so few people will read anyway.
Unless, of course, someone out there had unlimited mod points and was just trying to mod any censorship into the ground. But that could never happen, not at Slashdot.
switch ($decade) { case "the 50's": s/the Boogyman/Communist agitators/g; break; case "the 60's": s/the Boogyman/acid-eating hippies/g; break; case "the 70's": s/the Boogyman/disco/g; break; case "the 80's": s/the Boogyman/mutual assured destruction (and Grenada!)/g; break; case "the 90's": s/the Boogyman/evil hackers and George Michael/g; break; case "the 00's": s/the Boogyman/terrorists/g; break; }
Of course hardware and software companies won't use coercion to force people into TPM. They'll just stop selling everything else, citing "lack of demand." "There's just been no demand," Intel will say, "for a processor/mobo/whatever that doesn't support TPM, ever since Windows stopped supporting non-TPM platforms." Of course, months before, Bill Gates will have played the high morality card and announced that Windows would not longer run on non-TPM platforms; to allow that continue is to allow the continued spread of spyware and viruses, and Microsoft indignantly refuses to be any part of that!
See? It's not coercion. It's for security. It helps the economy. It thwarts terrorists. TPM gives flags to orphans if that's what it needs to do to get people on board.
Do what the GP did: ask your local university. We still have a nine-track drive around, although it hasn't been fired up in a few years. Lots of data from the state government, ACT test reports, etc., came on 9-track tapes until just five or six years ago, so lots of universities still have them around.
Let me quote myself: "My point is that none of the problems you point out are unsolved; we just need a comprehensive solution that includes all of the available technology."
Plus, the only users that would need to learn 5 different ways of installing, etc., are those who are running 5 different distros, i.e., emphatically not the average home user. The average home user has to learn one way.
My suggestion wasn't that home users need to use the command line. That's absurd. My suggestion (as I mentioned above, in my self-quote, and before, in my original post) was that all of the technology to make this happen exists. YaST lacks only elegant dependency handling. Yum and apt-get lack only a decent GUI. They all lack intuitive names. None of these things are difficult to implement, so, while installing is a sticking point of Linux, it is only very temporarily so. In other words: you, like the GP, are attacking a problem that is already being solved with great rapidity and fabulous success.
Linux configuration files: they are all in different formats, and follow different rules. (case sensitive? headings? comments? whitespace significance?)
And when does a desktop user ever need to edit a config file? (The answer is "never.")
Suppose I install Gnome as default and want to install KDE and use that as the default. How easy is that?
SuSE and Redhat both give me the choice every time I logon. It's in a nice, shiny menu. Maybe ucLinux or Slackware 1 or whatever backasswards distro you're using doesn't have that feature, but most do. Of course, you have to ask yourself: How many Joe Sixpack users out there want to switch desktop environments? They don't even know what the hell a desktop environment is. You're claiming that it's too difficult for beginners to do advanced things -- of course that's the case. It's also difficult for a Windows newbie to, say, change the letter their CD drive is mapped to. So what?
How many patches must you apply to SuSE right after install for all the security fixes? Dozens. Windows? Dozens.
So you're saying Linux is just as good as Windows? That's cool, I respect that.
Virtually every app available for Linux has a Windows counterpart. The reverse is also true.
Again, you seem to be arguing that Linux is just as good as Windows. I think you're a closeted "open-source fanatic" yourself. It's okay. You can come out to us.
And frankly, applying patches in Windows is easier than in Linux. Linux is pretty easy, but Windows is still ahead, I'd say. Argue all you want.
Huh? My Windows box and my SuSE boxes all automatically apply patches, but my SuSE boxes never need to be restarted afterward. I configured both to do automatic updates with a few clicks of the mouse; neither was particularly onerous.
And if you disagree, just know that most users don't know how to configure Linux like you do, and so your opinion doesn't matter to 98% of the computer users out there.
My favorite part of the article. A resounding ad hominem with a hilarious premise. Let me rephrase this sentence: "If you disagree with me, it's just because you're so smart, and smart people are irrelevant to this discussion." I'd call you an idiot, but you'd think I was complimenting you.
In all my conversations with Linux admins, I've never met one with a system configured where they didn't "su - root" to do admin stuff.
Really? You've never met anyone who used Ubuntu? Amazing.
Again, I'm not speaking of technical possibilities in Linux here, but what the trend is.
In other words, you're talking about culture, not software, and are by your own admission off topic.
I know Linux brags support recently for ACLs. My question: Where are they? They are so new (though Windows has had them since 1995) in Linux that no GUI interface I've seen supports setting/viewing them, and no command line "ls" will show those advanced permissions.
Finally! After 300 pages of ranting bullshit, we get to a real complaint. Yes, ACLs are new, although they're built in to the 2.6 kernel. I'm hoping that we'll start to see GUI support for them in newer distros.
Linux is cheaper if any of the following is true:[snip]
Or if you're a desktop user, at whom the original article was aimed anyway.
Windows is faster, and faster means cheaper!
Bwahahahahaha!
In summary, your whole article was hilarious, based on Linux distros that are apparently five years old, and more frequently attacked the Linux culture than the software. All in all, you had one good point, but even that one point is moot for the normal desktop user.
Basically, you, sir, are an ass.
When was the last time you had to choose between KDE and Gnome? I can't remember the last time I did. Maybe when I installed Redhat 8 or something.
I agree that installation needs some work, but a lot of that is getting users to understand that they don't need to go google for some crappy piece of freeware; they need instead to fire up YaST or whatever the equivalent is in your favorite distro and find it that way. (YaST is not a perfect tool; it's just the tool that I'm the most familiar with.) It's a different way to install, but not inherently any more difficult. Have you used yum or apt-get? I know, they're command line, but they resolve dependencies automatically. YaST does too, but it's a little more verbose.
My point is that none of the problems you point out are unsolved; we just need a comprehensive solution that includes all of the available technology.
My girlfriend runs Linux. She doesn't know what a shell is, nor does she care. (When was the last time you needed a shell for a (l)user-level function? Again, Redhat 8.)
Anyone who runs Ubuntu doesn't need to know what root is, either. You need to stop running Slackware 1, get with the program, and install one of the many polished, modern Linuxes with lots of promise to be viable competitors to Windows and Mac OS X. Seriously. Your post should be modded -1, Interesting Five Years Ago.
Slashdot Mirror
Gee, it sure is a good thing the end of WWII didn't precipitate five decades of near-constant petty warfare and general unrest! Whew! Dodged that bullet!
See, when you plan downtime, you plan more than you need. If you think a conversion will take six hours, you plan for twelve hours of downtime. The fact that the bank migrations always finish early means that nothing -- or very little -- goes wrong. That means that bank conversions -- which are far more common than you seem to believe -- go very smoothly and are much less fraught with danger than you'd make them out to be. (You'll note that the potential for data loss or corruption during the planning stages is numerically indistinguishable from zero.) I wouldn't call them "routine," but, with ample planning (as with any migration or conversion), as routine as such a thing can get. You presented bank migrations as something between brain surgery and bomb defusement on the delicacy scale; my point was simply that, given the rate of problems I've seen second-hand, it's nothing of the sort.
So I reiterate my initial point: given the degree of delicacy and potential data loss I see in the average bank migration, my bank can change OSes fifty-two times a year if they want. I don't care.
Clear now?
The HP DL 760 is not intended to compete with Sun's $750 (diskless) x2100.
I'm not sure where you got the idea that he's unaware of said complexity. You're inventing worse strawmen than the last guy.
That's what my roommate does. He works for a company that writes banking software, doing the conversions for banks moving from other systems, frequently running Solaris or AIX, sometimes running Linux, to theirs, which runs on Windows, which is a hell of a lot more difficult than switching between Unices.
Maybe you thought I didn't read the GP post, but I did. I didn't mention a damn thing about DNS or web servers; I was talking about banking software. Pull your head out and realize that knocking down a straw man doesn't make you big man on campus.
For the same reason that interesting stuff is getting modded down as troll. The mods are on crack.
That's a troll? Man, /. is getting weird. I miss the days when GNAA and goatse links were trolls. Seems like this guy (who has been modded up to +4 as of this post) was dead wrong on which way the ./ community leans when it comes to Sun.
HP DL 760
So in summary: I don't give a damn. As often as they want.
Dude, if you can't even figure out how to buy a service contract with your hardware, you sure as hell shouldn't be managing any system more powerful than a toaster.
Mmm, well designed server it looks like.
Sun has started down a long road of realization that their proprietary chips aren't worth the silicon they're printed on in a marketplace that values interoperability over nearly all else. HP realized this a while back, and have all but phased out their Alpha and PA-RISC lines.
That said, there is a place for non-x86 chips. HP has replaced most of their Alpha and PA-RISC lines with x86 chips, but some of the high-end boxes went to Itanium. IBM is still pushing POWER -- hard. I don't suspect that SPARC will die, but I do strongly suspect that, in 5 or 10 years, Solaris on x86 will be the standard and Solaris on SPARC will be the rare beast, still grinding away in a handful of data centers on Wall Street, right next to an HP box running Non-Stop OS.
However, depending on where Sun takes Solaris on x86, they could turn it into a huge market. They're already working on certifying most HP server hardware and, AFAIK, have their sights set on the server hardware of other major manufacturers as well. If Sun could make Solaris as easy to install anywhere as Linux or NetBSD (okay, so that's a bit of stretch), and if they keep giving it away for free, they could see sizable growth in their OS market.
The problem, of course, is that it's hard to make money off of a free OS, but, if you've been listening to Sun's web expos over the past year or so, you know that they label Redhat, somewhat strangely, as one of their primary competitors. Sun is obviously trying to stress their free OS; my question is whether or not they're trying to go too many ways at once. Sun provides not only a free OS, but also tons of enterprise software for it, plus hardware based on two different processing platforms. They're competing with Redhat, Novell, and HP all at once, and that's tough to do without stretching one's self too thin.
My prediction: if Sun is still a player in 10 years, it will be a very different Sun than we see today.
Oy! I understand that /.ers might not always RTFA, but can't you at least read the goddamn summary? It's a friggin' paragraph, it's not like you need to be in Mensa to understand it.
Unless, of course, someone out there had unlimited mod points and was just trying to mod any censorship into the ground. But that could never happen, not at Slashdot.
switch ($decade) {
case "the 50's":
s/the Boogyman/Communist agitators/g;
break;
case "the 60's":
s/the Boogyman/acid-eating hippies/g;
break;
case "the 70's":
s/the Boogyman/disco/g;
break;
case "the 80's":
s/the Boogyman/mutual assured destruction (and Grenada!)/g;
break;
case "the 90's":
s/the Boogyman/evil hackers and George Michael/g;
break;
case "the 00's":
s/the Boogyman/terrorists/g;
break;
}
See? It's not coercion. It's for security. It helps the economy. It thwarts terrorists. TPM gives flags to orphans if that's what it needs to do to get people on board.
Don't worry, it's available in PDF. I predict it's available over eDonkey or BitTorrent within the week.
Do what the GP did: ask your local university. We still have a nine-track drive around, although it hasn't been fired up in a few years. Lots of data from the state government, ACT test reports, etc., came on 9-track tapes until just five or six years ago, so lots of universities still have them around.
There you go. See, things are even better than I thought.
Plus, the only users that would need to learn 5 different ways of installing, etc., are those who are running 5 different distros, i.e., emphatically not the average home user. The average home user has to learn one way.
My suggestion wasn't that home users need to use the command line. That's absurd. My suggestion (as I mentioned above, in my self-quote, and before, in my original post) was that all of the technology to make this happen exists. YaST lacks only elegant dependency handling. Yum and apt-get lack only a decent GUI. They all lack intuitive names. None of these things are difficult to implement, so, while installing is a sticking point of Linux, it is only very temporarily so. In other words: you, like the GP, are attacking a problem that is already being solved with great rapidity and fabulous success.
In summary, your whole article was hilarious, based on Linux distros that are apparently five years old, and more frequently attacked the Linux culture than the software. All in all, you had one good point, but even that one point is moot for the normal desktop user. Basically, you, sir, are an ass.
I agree that installation needs some work, but a lot of that is getting users to understand that they don't need to go google for some crappy piece of freeware; they need instead to fire up YaST or whatever the equivalent is in your favorite distro and find it that way. (YaST is not a perfect tool; it's just the tool that I'm the most familiar with.) It's a different way to install, but not inherently any more difficult. Have you used yum or apt-get? I know, they're command line, but they resolve dependencies automatically. YaST does too, but it's a little more verbose.
My point is that none of the problems you point out are unsolved; we just need a comprehensive solution that includes all of the available technology.
My girlfriend runs Linux. She doesn't know what a shell is, nor does she care. (When was the last time you needed a shell for a (l)user-level function? Again, Redhat 8.)
Anyone who runs Ubuntu doesn't need to know what root is, either. You need to stop running Slackware 1, get with the program, and install one of the many polished, modern Linuxes with lots of promise to be viable competitors to Windows and Mac OS X. Seriously. Your post should be modded -1, Interesting Five Years Ago.