The first rule in data security is physical security as most boxes assume some level of trust for whomever can get into the same room.
For instance, one workable (done it) procedure to "fix" a lost root password on SCO, HP/UX, and AIX, is to crash the box, bring it up off different boot media, mount up the / partition, and edit/etc/passwd by hand.
This is why most serious data centers have locks on the doors.
Sark,
Ignoring your assertion that IE is best for the moment, why do you accept that MS has the right to make it easy to track someone elses' online activities?
Expert is the wrong term, although I do have some SCO certs as well as some IBM AIX and HP/UX certs in my file. They come in handy at re-negotiation time.
I agree that slapping a compatability layer on Linux isn't going to be easy, by the way, but in order to get the legacy users to upgrade into their product line, Caldera needs to think about it.
The trick question here is, what did Caldera really get in this deal? As I see it, UnixWare is pointless. It's not a direct upgrade for legacy OpenServer, so marginal sales there. Solaris for x86 is squeezing from the top and Linux et al [is|are] squeezing from the bottom. Caldera got the name and a nice OS, but no market to speak of.
The resellers frankly represent a whole business model was outdated years ago. I hate to sound negative here, but I think Caldera has made a mistake. I hope it isn't fatal...
I doubt, however, that the brains at Caldera would have made this move if there wasn't something of value left in SCO...perhaps we just don't see it yet?
While you're correct that MVS (OS/390) and whatever-the-old-name-of-the-OS-for-AS400s-was (OS/400)are going to be with us for some time, IBM does appear to be taking Linux seriously in that they've gotten it to run on the S/390 and AS/400 platforms, and what this release does is position it in their current moneymaker arena, the AIX server.
The upward-compatability is really nice for a growing company, though. "If we buy an RS/6000 43P, we can upgrade to an S/80, and then to an AS/400, and run it all on AIX/Linux". Linux becomes basically their hardware abstraction layer, which is, you must admit, a slick trick.
The others will be supported, but let's face it, most of what's really running on OS/390 is older code, and when the rewrite comes, being able to plunk cheap developers down in front of dirt cheap Linux beige boxes instead of more specialized MVS developers on terminals has a big cost advantage.
I'm not saying it's IBM's brightest move, but it does make a certain calculating sense, and it can't really hurt more than some of their other blunders. (OS/2 Warp edition?)
The problem I saw at two sites with existing OpenServer Unix servers is that the path from OpenServer to UnixWare was very rough.
The two are similar enough to let you get a false sense of confidence, but far enough apart that you could get into real trouble in the lower-level code or (my specialty) comms scripting.
Think of it as working in a shop that supports both the old SunOS and Solaris. Sure, they're both Unix from Sun...
Anyway, if Caldera is smart, they'll gin up a Linux that "emulates" OpenServer, then push that as the next logical step. Nice smooth progression, easy conversions, etc.
Oh, and don't forget - SCO has a legitimate (more or less) training and certification program. Being able to say "I know SCO, SCO says so" may not be as lucrative as a Microsoft certificate, but it is something the Linux world doesn't appear to have... yet.
More likely this is cheap (relatively) insurance. Could TransMeta afford to be tied up explaining the differences between emulating x86 and fab-ing x86 to a jury for two to five years?
If IBM weren't involved, I'd expect Intel to try something.
We're not talking about a 2-user or 5-user or a 50-user shop, though. We're talking several hundred nodes around the globe running heavy CAD apps on NT. I'm surprised he still has most of his hair, although it's getting thin on the sides.
As for Win2k vs Linux, I haven't used Win2k long enough to get a real opinion yet. It seems okay, but no great revolution. Nothing but bug fixes and some evolution and refinement, which while necessary are not earth-shattering and (imo) don't get a seperate version number. (95, 98, 00....)
I hate to say it, but it looks like RMS may have been right - Linux is certainly developing at a much faster pace than MS. The future doesn't look bright unless MS can innovate or effectively compete, and how do you fight something that's free?
If you define using it right as putting it all back in the cardboard box, then using it as a paperweight, then maybe.
Beyond that, I can tell you a shop I know someone at is rolling out NT 4 with an HP-ish professional grade engineering modeler to engineers used to Unix (HP/UX) workstations.
It's actually quite simple - but expensive. IBM, for instance, will sell you whatever talent you want.
Worked for a shop once that did a DB2-driven kiosk app on top of NT on IBM workstations using whatever the hot GUI display technology was at the time. IBM touchscreens, IBM custom membrane keyboards, IBM designed DB, they did it all.
The thing looked fine when it worked, but after a day or two started bogging down badly, and they had hardware problems you would not believe.
Everything was the responsibility of this one guy at IBM, and I would not have taken his job for all Gates' money. That he lived through a coronary during this project may tell you something of his stress level. We eventually got things up enough to roll it out, and then started having problems with the IBM-provided credit card authorisation software. That's about where it was when I quit, and I doubt it's gotten further. It was a genuine IBM-designed IBM-implemented disaster...but yes, they'll sell you an external resource whose job it will be to manage such.
I support a remote data center consisting of mostly HP kit connected to EMC disk arrays.
When we have disk problems that I can't work out remotely (i.e. EMC claims their rig is fine, but the HP kit can't see disks and HP are claiming it's fine) I'll call out hardware techs for both, once they're at the site, conference 'em both in and say "Figure it out".
Usually works pretty well, but it can take some doing to get 'em there as neither claim to have a problem, so don't want to dispatch a tech. Waving a P.O. number usually helps...
Shop I worked for put SAP on Oracle on AIX on RS/6000 S70s. Know where support fell apart? IBM.
The S70 is a beauty of a box, and the hardware support engineer really knew his way around inside it. The real problem is the VAR brain drain. If you're hired and trained on AIX by IBM, odds are the resellers will agressively headhunt you. The result is, the odds of getting someone on the support line who actually knows anything is pretty low. I ended up figuring a lot more out from my System V background and reading the manuals while on hold for support than I did from support.
Linux serving Samba vs. NT serving SMB (i.e. a departmental file share server) the story makes sense, especially since Linux/Samba blows NT/SMB out of the water. You are correct, user desktops are not there...yet.
I think you're right about ability to sue - technically, a lie or falsification cost her the job, and that's going to follow her every time she changes companies.
"X didn't hire you why?".
I would look to sue against lost wages and possibly earning potential.
Of course, if said friend really wants this job, then as another poster has said, "prove" that she didn't do it and hope for the best.
Hey, if it doesn't work out, then you've got CLEAR proof of lost wages...
I have to write up "ops procs" (Operations Procedures) for anything I assign to our Ops staff, and usually they skip the "definition of terms" and go right to the "type this, type that".
I've started putting restaraunt reviews in the definitions, just to see who's paying attention.
Microsoft wrote Xenix. This was an attempt to compete with: (a) DR's Concurrent DOS (b) Unix (c) CP/M depending on whom you believe.
My guess is, one of the early Microserfs thought Unix rocked (rightly so) and decided to do what Linus did...just within a corporate structure.
They couldn't make enough money at it, and it conflicted with MS-DOS, so out it went. The result was SCO, which later bought "Unix" from Novell, so both MS and Novell own big (>5%) stakes.
People on/. seem to treat "business people" as if they were an alien race.
There's a good reason for this - business people are not engineers or coders or admins, they don't think the way we think.
Red Hat is a publically traded stock that even your grandmother has heard of.
Being publicly traded doesn't mean jack. Also, just because the Men In Suits have heard of a product doesn't mean they see the value in it.
The "Men with Big Chequebooks" see SCO the same way you do - a has-been looking for a reason to survive.
Your most foolish statement - the Men In Suits see SCO as someone they've written a cheque to before, someone they may have a support contract with. In other words, a known good (well, adequate) commodity. To a Suit Man, this is a Good ThingTM.
Please try to make your way from your dorm room to an accounting or management class next semester, okay?
Comments in code are like a time capsule, from one programmer to another. They communicate what the code does, but also the mood of the author.
My first coding job was cleaning and adding "improvements" to some anchient and nasty C code. Some of the comments described the previous programmers' marital problems, opinions on co-workers, etc. It made a sucky job more bearable.
So your cat has good taste. Good for you. Now, what OS does Brutus prefer?
I haven't slipped out to see this film, and given the reviews here, I doubt I will. Not worth the hassle of losing my voice meowing outside the window at 2:00am to get let back in after.
Slashdot Mirror
The real question is whether it's undocumented features or beauraucratic inertia that's holding them back...
Meow
The first rule in data security is physical security as most boxes assume some level of trust for whomever can get into the same room.
/etc/passwd by hand.
For instance, one workable (done it) procedure to "fix" a lost root password on SCO, HP/UX, and AIX, is to crash the box, bring it up off different boot media, mount up the / partition, and edit
This is why most serious data centers have locks on the doors.
Meow
Death of MS-OS is unlikely.
MS-OS releasing something completely different (i.e. MS/GNU/Linux) and loading it up with all kinds of proprietary "MS-Win compatability tools"....
Meow
Sark,
Ignoring your assertion that IE is best for the moment, why do you accept that MS has the right to make it easy to track someone elses' online activities?
Meow.
Yep, that's pretty much the "Big Media" position. Eloquently put.
Meow.
Expert is the wrong term, although I do have some SCO certs as well as some IBM AIX and HP/UX certs in my file. They come in handy at re-negotiation time.
I agree that slapping a compatability layer on Linux isn't going to be easy, by the way, but in order to get the legacy users to upgrade into their product line, Caldera needs to think about it.
The trick question here is, what did Caldera really get in this deal? As I see it, UnixWare is pointless. It's not a direct upgrade for legacy OpenServer, so marginal sales there. Solaris for x86 is squeezing from the top and Linux et al [is|are] squeezing from the bottom. Caldera got the name and a nice OS, but no market to speak of.
The resellers frankly represent a whole business model was outdated years ago. I hate to sound negative here, but I think Caldera has made a mistake. I hope it isn't fatal...
I doubt, however, that the brains at Caldera would have made this move if there wasn't something of value left in SCO...perhaps we just don't see it yet?
Meow
While you're correct that MVS (OS/390) and whatever-the-old-name-of-the-OS-for-AS400s-was (OS/400)are going to be with us for some time, IBM does appear to be taking Linux seriously in that they've gotten it to run on the S/390 and AS/400 platforms, and what this release does is position it in their current moneymaker arena, the AIX server.
The upward-compatability is really nice for a growing company, though. "If we buy an RS/6000 43P, we can upgrade to an S/80, and then to an AS/400, and run it all on AIX/Linux". Linux becomes basically their hardware abstraction layer, which is, you must admit, a slick trick.
The others will be supported, but let's face it, most of what's really running on OS/390 is older code, and when the rewrite comes, being able to plunk cheap developers down in front of dirt cheap Linux beige boxes instead of more specialized MVS developers on terminals has a big cost advantage.
I'm not saying it's IBM's brightest move, but it does make a certain calculating sense, and it can't really hurt more than some of their other blunders. (OS/2 Warp edition?)
Meow.
The problem I saw at two sites with existing OpenServer Unix servers is that the path from OpenServer to UnixWare was very rough.
... yet.
The two are similar enough to let you get a false sense of confidence, but far enough apart that you could get into real trouble in the lower-level code or (my specialty) comms scripting.
Think of it as working in a shop that supports both the old SunOS and Solaris. Sure, they're both Unix from Sun...
Anyway, if Caldera is smart, they'll gin up a Linux that "emulates" OpenServer, then push that as the next logical step. Nice smooth progression, easy conversions, etc.
Oh, and don't forget - SCO has a legitimate (more or less) training and certification program. Being able to say "I know SCO, SCO says so" may not be as lucrative as a Microsoft certificate, but it is something the Linux world doesn't appear to have
Meow
More likely this is cheap (relatively) insurance. Could TransMeta afford to be tied up explaining the differences between emulating x86 and fab-ing x86 to a jury for two to five years?
If IBM weren't involved, I'd expect Intel to try something.
Meow.
Point noted.
We're not talking about a 2-user or 5-user or a 50-user shop, though. We're talking several hundred nodes around the globe running heavy CAD apps on NT. I'm surprised he still has most of his hair, although it's getting thin on the sides.
As for Win2k vs Linux, I haven't used Win2k long enough to get a real opinion yet. It seems okay, but no great revolution. Nothing but bug fixes and some evolution and refinement, which while necessary are not earth-shattering and (imo) don't get a seperate version number. (95, 98, 00....)
I hate to say it, but it looks like RMS may have been right - Linux is certainly developing at a much faster pace than MS. The future doesn't look bright unless MS can innovate or effectively compete, and how do you fight something that's free?
Meow
Windows doesn't crash if you use it right?
If you define using it right as putting it all back in the cardboard box, then using it as a paperweight, then maybe.
Beyond that, I can tell you a shop I know someone at is rolling out NT 4 with an HP-ish professional grade engineering modeler to engineers used to Unix (HP/UX) workstations.
They are not having an easy go of it.
Meow
It's actually quite simple - but expensive.
IBM, for instance, will sell you whatever talent you want.
Worked for a shop once that did a DB2-driven kiosk app on top of NT on IBM workstations using whatever the hot GUI display technology was at the time.
IBM touchscreens, IBM custom membrane keyboards, IBM designed DB, they did it all.
The thing looked fine when it worked, but after a day or two started bogging down badly, and they had hardware problems you would not believe.
Everything was the responsibility of this one guy at IBM, and I would not have taken his job for all Gates' money. That he lived through a coronary during this project may tell you something of his stress level.
We eventually got things up enough to roll it out, and then started having problems with the IBM-provided credit card authorisation software.
That's about where it was when I quit, and I doubt it's gotten further. It was a genuine IBM-designed IBM-implemented disaster...but yes, they'll sell you an external resource whose job it will be to manage such.
Meow
I support a remote data center consisting of mostly HP kit connected to EMC disk arrays.
When we have disk problems that I can't work out remotely (i.e. EMC claims their rig is fine, but the HP kit can't see disks and HP are claiming it's fine) I'll call out hardware techs for both, once they're at the site, conference 'em both in and say "Figure it out".
Usually works pretty well, but it can take some doing to get 'em there as neither claim to have a problem, so don't want to dispatch a tech. Waving a P.O. number usually helps...
Meow
Shop I worked for put SAP on Oracle on AIX on RS/6000 S70s. Know where support fell apart?
IBM.
The S70 is a beauty of a box, and the hardware support engineer really knew his way around inside it.
The real problem is the VAR brain drain. If you're hired and trained on AIX by IBM, odds are the resellers will agressively headhunt you. The result is, the odds of getting someone on the support line who actually knows anything is pretty low.
I ended up figuring a lot more out from my System V background and reading the manuals while on hold for support than I did from support.
Meow
Linux serving Samba vs. NT serving SMB (i.e. a departmental file share server) the story makes sense, especially since Linux/Samba blows NT/SMB out of the water. You are correct, user desktops are not there...yet.
Meow.
But can it hold it?
I could always take Asia or Europe, but couldn't keep 'em.
I think you're right about ability to sue - technically, a lie or falsification cost her the job, and that's going to follow her every time she changes companies.
"X didn't hire you why?".
I would look to sue against lost wages and possibly earning potential.
Of course, if said friend really wants this job, then as another poster has said, "prove" that she didn't do it and hope for the best.
Hey, if it doesn't work out, then you've got CLEAR proof of lost wages...
Meow.
IIRC, MMDF was the "default" prior to OpenServer 5.0.5, after that it became Sendmail.
I configured both, and found MMDF to have marginally more "obvious" configuration, but significantly less functionality.
Meow
Good! Then I'm not the only one who does this!
I have to write up "ops procs" (Operations Procedures) for anything I assign to our Ops staff, and usually they skip the "definition of terms" and go right to the "type this, type that".
I've started putting restaraunt reviews in the definitions, just to see who's paying attention.
Meow
HP Intel-based workstation runs Linux - FAQ at 11
Meow!
Microsoft wrote Xenix.
This was an attempt to compete with:
(a) DR's Concurrent DOS
(b) Unix
(c) CP/M
depending on whom you believe.
My guess is, one of the early Microserfs thought Unix rocked (rightly so) and decided to do what Linus did...just within a corporate structure.
They couldn't make enough money at it, and it conflicted with MS-DOS, so out it went. The result was SCO, which later bought "Unix" from Novell, so both MS and Novell own big (>5%) stakes.
Meow
People on /. seem to treat "business people" as if they were an alien race.
There's a good reason for this - business people are not engineers or coders or admins, they don't think the way we think.
Red Hat is a publically traded stock that even your grandmother has heard of.
Being publicly traded doesn't mean jack. Also, just because the Men In Suits have heard of a product doesn't mean they see the value in it.
The "Men with Big Chequebooks" see SCO the same way you do - a has-been looking for a reason to survive.
Your most foolish statement - the Men In Suits see SCO as someone they've written a cheque to before, someone they may have a support contract with. In other words, a known good (well, adequate) commodity. To a Suit Man, this is a Good Thing TM.
Please try to make your way from your dorm room to an accounting or management class next semester, okay?
Meow
Kinda like PC-DOS, MS-DOS, and DR-DOS (although late) went after one another in the '80s...
Begs the question - who will be the Microsoft of Linux distros?
Meow!
Comments in code are like a time capsule, from one programmer to another. They communicate what the code does, but also the mood of the author.
My first coding job was cleaning and adding "improvements" to some anchient and nasty C code.
Some of the comments described the previous programmers' marital problems, opinions on co-workers, etc.
It made a sucky job more bearable.
Meow!
So your cat has good taste. Good for you.
Now, what OS does Brutus prefer?
I haven't slipped out to see this film, and given the reviews here, I doubt I will. Not worth the hassle of losing my voice meowing outside the window at 2:00am to get let back in after.
Meow!