SCO UnixWare 7.1.3 Review

JigSaw writes "Despite news about SCO being all about the lawsuit, they still sell OS products and they have a presence in the server market. UnixWare is one of these OS products. Tony Bourke reviewed its latest version, 7.1.3, and even includes benchmarks among other tests. Tony concludes that 'the lack of commercial applications and user community, the difficulty with open source applications, the SCO litigation, and the high price are all marks against UnixWare. There are just very few reasons to adopt UnixWare as your platform, and plenty of reasons to adopt (or migrate to) other platforms.'"

it was an objective review

[Anonymous Coward]

I thought the author did fairly well at remaining objective and testing the product without allowing company ethics cloud his review

Expensive and sparsely featured...

[cheesybagel]

Why would anyone choose it over Linux of FreeBSD is over me.

A prediction...

[Fnkmaster]

I think their UNIX business will get spun off after the lawsuit business clears up and the company goes bust. The Unixware product will no longer be marketable under the "SCO" name, since the brand will be indelibly tarnished in the IT world as part of a hostile, litigious organization that tried to extort money from companies, big and small, for work that they had no rights to, and for what essentially amounts to a massive pump-n-dump scheme.

This may hurt them the most...

[wrinkledshirt]

Of all the bad PR that they've generated for themselves, a bad product may hurt them the most. Now, they open themselves up to the counter-attack that they're an untalented software company looking for a quick buck, with the product being proof of their lack of talent. It's an oversimplification, sure, but one they pretty richly deserve.

expensive crap

[potpie]

Unixware proves that sometimes, (an increasingly small number of) people buy things based on price alone. There is no reason to use such an expensive, restrictive OS when the makers of that OS have to use ideas from their biggest competitor to improve it, when that competitor is a free (in all meanings) OS.

Let's not get into the specific advantages, because nobody has that large an attention span.

Re:Unfortunate that the test system wasn't newer

[Anonymous Coward]

You can be sure that it doesn't. Hell, Linux barely supports most of that sutff.

Unixware customers (if in fact they exist) are going to be very conservative with hardware -- they will buy from an integrator that uses systems/parts that are listed on the HCL. Period. Gamer stuff like nForce boards is irrelevant.

Re:surprise surprise

[Anonymous Coward]

Put it under any serious server load on a multiproc box and it probably would kick the shits right out of Linux 2.4.

Unixware is basically the same thing as Solaris. Even Linus et al admit they aren't at that level yet.

Linux is only "better" because of broader app & hardware support. Even the price difference wouldn't matter that much if you are buying Oracle licences or something.

can I get a liscense with that?

[UltraSkuzzi]

I bought a copy of SCO UnixWare in the pre Darl days. Back in that day, you could get free 'educational' licenses for nonprofit uses. Too bad they don't offer free 'linux' licenses for schools & colleges. Yeah I guess, they 'changed there minds'. UnixWare 7 wasn't a bad OS, but I believe the review was on target when he said the technology it's based on is past its prime. And you gotta love how you need a license for everything from SMP to critical security updates.

Re:Hmmm, a link to Microsoft?

[jhunsake]

He is untarring the tcsh distribution so he could compile it. tcsh includes support for Windows.

Re:Unfortunate that the test system wasn't newer

[Anonymous Coward]

They will support them all since all they do is strip the copyright from the Linux source code and sell it binary only.

Fuck you Darl.

Re:Anyone else find this funny?

[Anonymous Coward]

Hillarious!

Even better with a clickable link:

http://uptime.netcraft.com/up/graph/?host=sco.com [netcraft.com]

Re:wtf???

[Elwood P Dowd]

We're not talking about desktop users here. For the people that may want to use either UnixWare or Linux, Linux supports those features while UnixWare does not. Less sophisticated users should stay far, far from either Linux or UnixWare, "The Unix That Crashes (TM)". Scaring up a custom kernel is not exactly rocket science.

Re:Unfortunate that the test system wasn't newer

[Gojira Shipi-Taro]

Application vendors are dropping support for SCO right and left, so really, the level of hardware support is irrelevant.

I find it hard to believe that any company that has made the dire mistake of tying themselves so closely to SCO as a platform would not be actively investigating any possible option to remove themselves from any involvement at all with a clearly doomed company.

Their product is worthless, and their user base is so miniscule as to make it counter productive to expend the cash required to qualify product against SCO.

And the more that happens, the worse it will get for those who persist.

What good is an OS distribution when no one makes applications for it anymore, and those that did DROP support for it completely, because it's cheaper to lose a miniscule number of customers than to spend time and money supporting the OS they use?

Re:surprise surprise

[Geek of Tech]

>> Linux is only "better" because of broader app & hardware support.

Uh... isn't that the same reason all the Microsoft zealots use for saying Windows is better?

Benchmarks?

[molo]

The only benchmarks run were comparing OpenSSL computation in native UnixWare mode versus Linux Kernel Personality (LKP) mode. This is an extremely poor test and shows that the reviewer doesn't know what he's talking about.

LKP is basicly system call emulation like that which is available in FreeBSD. This has NOTHING to do with pure user-space number crunching required of crypto computations! This kind of test would only show the most eggregrarious scheduling or interrupt handler errors in providing the LKP functionality. This wouldn't (shouldn't?) even show up any compiler differences between UnixWare's cc and GCC since OpenSSL is heavily assembly optimzed on x86.

These numbers arn't even compared to running under a real Linux kernel, which would be the most logical course of action given the reviewer's incomplete understanding.

But regardless, with comments like the following, it becomes painfully obvious the reviewer knows little about this:

The Linux kernel version number piqued my interest, because of the recent kernel vulnerability responsible for the compromise of some Debian project servers. I'm not sure if the same kernel exploit would work in the LKP, but it'd be an interesting test.

If anything, benchmarking system calls should have been done. Something along the lines of these tests [bulk.fefe.de].

The reviewer makes his bias very plain with passages such as:

I want to be as objective as possible, but I'd be a fool to think such a review could possibly avoid the controversy and raw emotions surrounding the company offering the product I've chosen to evaluate.

This combined with the lack of objective and useful benchmarks makes this article little more than a piece of cheerleading propoganda.

-molo

Re:surprise surprise

[be-fan]

Unixware is basically the same thing as Solaris.
>>>>>>>>>>>
No, it isn't. They're both derived from SVR4, but all the performance insanity that Sun put into Solaris went in *after* the split.

Re:wtf???

[Anonymous Coward]

It supports SATA and IDE RAID, but the drivers aren't there for a lot of controllers. You could say that's hardly support at all, but by that logic you could also say because Linux doesn't support Brand X video card, Linux doesn't support graphics.

There's a difference between driver support and feature support. Linux supports these features. Drivers, as usual, depend on vendor specs, vendor support, and ease of reverse-engineering.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Expensive and sparsely featured...

[ChocolateCheeseCake]

Thats the thing, they don't choose it. Most of the time companies only keep with something this tragic if they need the legacy support, however, with that being said, Solaris does support *some* UnixWare binaries, if they improve their support to a decent level and offer "competitive upgrades", I am sure every man and their dog will move.

SCO has to stratergy for the long term. Their viability as an organisation is dead regardless of the outcome of this whole SCO vs. IBM trial. The fact is that the so-called "features" they accuse IBM of adding have existed in the kernel since the very early 2.4 betas. If I was the judge I would question why it has taken SCO 2-3 years to come out and say something? are they so desperate for cash, they're willing to sue all and sundry in a vein of hope of leaving with some sort of "legacy"?

I'm not Linux fan, heck, I run FreeBSD, MacOS X and Solaris, however, if one wants to compete with Linux, there must be a compelling reason. Solaris 10 has some outstanding features that will push it ahead of Linux. That is how you compete, by creating a better product, not going around and threating companies like SCO is doing right now.

Swedish Army

[W2k]

Last time I served in it, the Armed Forces of Sweden [www.mil.se] were still running SCO Unix for a lot of communications control computers. The systems were very buggy and would often crash. When I left they were just starting to migrate over to (customized) FreeBSD boxes and Windows NT. Now, knowing the Swedish army, I know they are NOT an organization that changes it ways unless it desperately has to (despite what their PR keeps saying). So if they're dropping SCO .. well .. I used their old systems myself, so I know pretty well how much they suck. It's all over for SCO. When all the legal bullshit is done and over with, there'll be nothing but bones left, and maybe the world will be rid of the horror that is UnixWare.

Re:Unfortunate that the test system wasn't newer

[onomatomania]

That's easy to say, but if you have critical infrastructure built around SCO it's not like you can just wake up one day and say "Hmmm, this doesn't look good, how about we abandon all those production servers and build something completely different." In business, things that work and are supported don't get touched without good reason, especially if megabucks have been spent getting to that point. It doesn't matter if SCO doesn't have shit for features or doesn't support the latest doodads. It's in production in a number of places and you can't just yank the rug out from under a business like that.

It's one thing to denounce SCO for being the assholes that they are, but it's another completely different thing to actually move away from something that critical without a LOT of planning and testing. Sure, you get started on that as soon as possible, but it takes time. YOu can't just say "SCO's irrelevant now" because to some businesses, it's very relevant -- for better or worse.

Re:Unfortunate that the test system wasn't newer

[Anonymous Coward]

Err..
With that reasoning your running into the concept of sunk cost fallacy.

Investing more money in SCO is throwing good money after bad.

It doesn't matter if you spent 25,000 dollars on a Linux infrastucture or a 250,000 SCO infrastructure. That money is gone and your not ever going to get any of it back.

It's like a car. I bought a big car and the engine blew. So I replaced it. 2500 dollars. Then the rear frame was rotted out from rust so I fixed that. Another 1500 buckaroos. Now the transmission blew and the brakes need replacing. 3500 Dollars will be needed to fix that.

Now is it smart to say:

"I spent 4000 dollars on that car. I don't want to waste that money. So I should spend 3500 dollars to save it."

OR

"That car is a POS. I shouldn't of spent the money on it, I shouldn't of got it in the first place. It was a mistake and I learn from mistakes so I should spend 5000 on a decent small car to replace it."

Now apply that to SCO:

"My company spent 3,570,000 dollars on a infrastructure built around SCO. This infrasture is worthless and can be outperformed by a rival company's that cost a third of that. I should spend another 750,000 on new servers so that I can get it working so I don't lose the 3,750,000 dollars that I spent in the first place."

or

"I am wasting huge amount of money on crappy SCO software. Application support is more and more expensive and SCO has showed incompitance in not patching a server that got attacked by a 4 year old DDOS exploit. I can't compete with my competaters and am losing money. I need to spend 300,000 dollars on a decent OS. (Linux)

So what would be the company that you would like to invest in? One that blindly thows it's money at problems or one that actually learns and grows in capabilities?

Re:Wake up one day?

[onomatomania]

I'm not saying that it represents necessarily good decision making on the part of those companies... But I'm just pointing out that it's all too common to have some random server running some random application (that's probably itself very old) that's crutial to the business. Nobody in the company has ever tried it with any other platform, nobody knows if it would work, no one knows how long it would take to switch formats or port the app, nobody knows how long it would be down while all this is going on, etc. When you have a situation like this that's crutial to the business functions of the company and it's working and supported, it's going to be an uphill battle to convince anyone to change, ESPECIALLY to commodity and/or "community supported" stuff.

Please, don't take this as me trying to justify SCO's crapware in any respect. I'm just trying to point out that if you spend a lot of time in open source circles it's very easy to get this skewed version of things in which it's inexplicable why any company wouldn't have burned every last piece of SCO media and torn up every support contract after months of this lawsuit garbage and years of crappy software that's going nowhere. You'll find that businesses often have tons of random legacy junk sitting around that's still useful, and to keep it running it makes more sense from a business standpoint to keep paying SCO for support contracts or upgrades, regardless of the merits of SCO's software. SCO knows this, and they have to play into it if they want to survive... (Or at least, a semi-sane SCO before all this lawsuit crap. Now they've pretty much made it impossible to survive post-lawsuit.)

It's kind of like the tale ('Signs'?) where the car runs over the man and pins him against a tree or wall or something, holding his innards in place. You know that his game is up sooner or later but you also know that moving the car is going to make a huge mess with his guts oozing out everywhere...so it's best to just keep things as they are for as long as possible until at least the EMT arrives and he has a slight chance of surviving.

Re:Unfortunate that the test system wasn't newer

[Anonymous Coward]

especially if megabucks have been spent getting to that point.

Actually, one of the lessons of capitalist economic theory is "sunk costs are sunk costs". If you've made what turns out to be a mistake in the past, if you let it influence your future behaviour (as a naive western human in CYA mode tends to) in a capitalist system, you'll eventually lose. This isn't hearsay, this is a result that drops out of economic modelling. All the crap that corporations spew to justify fundamentally anti-free-market intellectual "property" laws about having a "right" to recoup investments/R&D costs is just that: crap, and dangerous too as it hands rather nasty people tools to silence dissenters and crush competitors all legit and above-board, like. Not good.

Re:Wake up one day?

[Rinikusu]

Exactly. I know quite a few restaurant chains that use SCO UnixWare machines that are already bought, paid for, and applications developed for, and for the most part, they work. Just up and scrapping SCO would mean serious downtime measured in the tens of thousands of dollars in lost revenue because their computerized ordering/order-fulfillment systems are down and manual entry (which has to be entered later) is just slower than normal. These are also companies who probably PAID SCO to come up with their custom solutions, so they're reluctant to have to fork that money out yet again to someone else willing to port/reimplement the system for Linux or what not. (Much of the software is not owned by the company running it).

A previous company I worked for had 2 SCO servers for only 1 reason: the courier dispatch/order entry/database software they needed only ran on SCO. They looked for other prebuilt solutions, none of which were suitable (each package lacked one or two critical features that the current package provided). So, while they would have loved to move to an all MS shop (or even move to Linux), they were limited by their choice in software. Sure, they *could* probably spend a couple hundred thousand dollars for a few programmers to come up with a custom solution, but what they have works now and they have exactly *zero* programmers on the dole. Small companies don't have the resources to fill the gaps, so to speak..

Hrm, speaking of which, that gives me an idea...