Slashdot Mirror
An Objective Review of UnixWare 7.1.4
Roblimo writes "Yes, SCO is evil and all that, but in between lawsuits it still puts out a product called UnixWare. NewsForge decided to review the latest version -- 7.1.4 -- just like we would any other Unix-based operating system. To ensure impartiality, we hired respected freelancer Logan G. Harbaugh, who wrote: 'On the server side, UnixWare Enterprise edition is more expensive for 150 users than either Windows 2003 Server Datacenter Edition, any of the Enterprise Linux distributions, or Solaris, with fewer available applications, fewer drivers for recent HBAs and other new hardware, and no currently available 64-bit version for either Opteron or Itanium processors.'"
Why? Just because someone casts Windows in a favorable light, they automatically lose their objectivity?
Rumors are that many lines of source code are in common with Linux; and so far all the lines identified actually belonged in Linux.
The only natural conclusion is that these guys stole them.
SCO, Please pay Linus $699,000 for every copy you sell.
...when you replace developers with lawyers.
I pity the foo that isn't metasyntactic
The funny thing is, for as much as our friends at SCO are threatened by OpenSource, OS is the only way that they can compete with larger entities like Sun and HP. Look at how many of the above list of new "features" are simply OSS ports. Think of how much work it would have been for SCO, and their handful of engineers to recreate these ports from scratch.
After looking at these points, why are we to assume that SCO is losing money because of Linux infringing on their IP? Isn't it more likely that SCO has just lost touch with the market, and has been passed up by better competitors?
Overrated / Underrated : Moderation
That SCO is proclaiming the GPL to be the world's greatest evil while still shipping a billion GPL applications in the box.
Or maybe it just thinks it "owns" the applications as well?
And yes, I had lots of clients that used Xenix. And after a LOT of pain, they figured out that SunOS on a $15k sun was cheaper than Xenix on a $3000 386/25.
My friend still hates me for making him setup an early Unixware and it's NIS/YP "implementation" (rsh to master, copy files over, merge with local ones, done. That's like NIS, right mr customer?)
We tossed it for being grossly unsecure, even on a trusted LAN; slow and bad.
Oh, and the switch was set to Evil (but we didn't know it then).
I was in a unixware shop many years ago, and the best thing I like about it was a piece of software called Merge. But a couple years ago, Win4lin, came out for Linux. This was back in the multi-cpu 486 days, made a great call center server with a hundred operators on it. Other than than the printer queue's fscking up, it was stable. But I was already running BSD for any server I was tasked to engineer.
Today, I dont really see a reason to use unixware. The software is all GPL'ed software you can download on most platforms, and Solaris and Linux have better support.
Just my 2cents.
Dear SCO,
It was a better effort than we expected of you.
However, it still sucks.
Respectfully,
NewsForge
Well, the actual introduction to the review reads:
UnixWare 7.1.4 is the latest in a long line of Unix releases from The SCO Group. It is a stable and mature Unix, with a variety of basic servers included, such as the Apache Web server and Squid, and is available in both single-user desktop-oriented versions and server versions. It has reasonable support for hardware, good documentation, and a nice integrated management utility that offers unified administration of the OS, hardware, and servers. Performance as a server platform is good, supporting a number of TCP sessions and Web server users, and file transfer performance is competitive with Linux and Windows platforms. However, as a desktop OS or file/print server, UnixWare is hard to recommend over competitors.
And the actual conclusion:
UnixWare 7.1.4 offers some high quality Unix features including OS stability and security, disk replication, a decent GUI management package, Windows emulation, good documentation, and a reasonable suite of server applications. However, the relatively high prices for adding multiple users and CPUs, high cost of the support package, and relative dearth of available software since the LKP package was removed make UnixWare hard to justify as a file/print or mail server, or desktop OS. It would make a good Web server or application server.
Doesn't sound quite as bad as the slasdot summary, does it?
... If you still want to make a profit, you have to charge more for your software when there's fewer people willing to buy it.
Sounds pretty objective to me. The author's point is simply that the price seems out of whack with reality, and objectively that is absolutely correct. SCO needs to make a strong case as to what you're getting for the extra money when compared to Windows, Solaris, Linux (Enterprise Linux, with paid support), or any other OS you might choose. Nothing I've heard from SCO or seen for myself leads me to believe the price is justified compared to the competition. The support isn't significantly better. The reliability isn't significantly better. The number of available applications is much worse. Etc. And then you have the whole "will this company be around in a couple years" problem with SCO. Even ignoring the lawsuit fiascos, they've had a number of other business-related problems as of late (with Baystar, etc) that would pretty much ensure I'd never choose an SCO solution for a new project, even if it were cheaper instead of more
Ok, nice and all.. lets be objective.
Sorry. As the systems/network engineer here, I get a fair amount of say in what goes and what doesn't, and even a bone-headed PHB (and I've got 2 out of three directors who fit that mould.. and I can say that here because I'm changing jobs anyway in a couple of months) can see that anything that makes as much noise as SCO is not a long term bet.
Short of it: Doesn't matter if Unixware is great or crap if its not a cast-iron guarantee that the company will be around in 3 years to support the platform.
Norman Cook's Ode to Sl
From reading the comment, I'm not so sure
So, exactly which part of "more expensive" and "fewer drivers" are not 100% quantifiable and objectively measured criteria? This is the measure of an objective interviewer, rather than comments like "The install process was confusing" and "The GUI sucked!"
I've used Xenix, Open Desktop, Open Server, and fairly recent versions of SCO UNIXware. I like Xenix the best of the lot.
The original UNIXware when Novell had it was pretty good, actually. Not like your typical commercial UNIX, because it didn't have nearly as much BSD influence as most of the survivors, but it did the System V thing as well as anything I've used. Don't expect it to be real happy in a BSD/Sun/Linux environment, but off by itself or surrounded by Windows or Netware boxes it was pretty solid. And, after all, that's what they sold it for.
The version I used after it had been in SCO's hands had an awful lot of Open Server in it, and it suffered from the transplant. The biggest problem was that the system just had too many different subsystems and components each with their own configuration interfaces all hidden behind their clumsy (but not much worse than other CDE-ish front ends I've used) GUI configuration tools. The result was that when things went wrong it was terribly difficult to diagnose.
This isn't something that you're likely to notice until you'd actually been using it in production a while, unless you had the bad (or is that good) luck to step on a crack in the initial install.
Back when it was Xenix, particularly the early versions, it was a lot more coherent and internally consistent. They really did start out with a pretty good system for the market they were selling into.
Given SCO's behavior, really, I've got to wonder who'd purchase this. I'd think pretty much any Unix-leaning admin or CIO knows what SCO's been up to this past couple years, and will summarilly dismiss it whether it's good or not. Plus no Windows-leaning admin or CIO would buy it in the first place.
So who is the target market?
#DeleteChrome
-I hate Open Source.
-But you're using it in your own products!
-The best there is!
-But you just said you hated it!
-But.. the you who.. I... It's... differeee.... (head explodes)
Can you hear me, Major Tom? I'm not the man they think I am at home...
Believe it or not, we are still nursing a few old SCO Openserver 5.0.x boxes along. Recently, I tried to purchase a SCO 5.0.8 because, I believe SCO is going to go belly up soon, and I wanted any last drivers they may have compiled into their O/S... I had to order the media and license separately. The SCO 5.0.8 media showed up, but the license has been backordered for about a month. It's really wierd that a piece of paper containing a license key could ever be on backorder. Maybe SCO fired their printer after all their NEW Linux license keys didn't sell.
I just wondered if anyone else has experience has tried to purchase any SCO product lately and experienced anything similar. Also, if anyone has any unused SCO 5.0.8 licenses they want to sell, please let me know. We are going are best to move off of SCO, but unfortunately some of the old applications just won't DIE easily.
Did you READ the article?
... of course, I admit to being biased. That he still decides it's a bad choice merely echos what most of the market has already decided, so it's hard to call that biased.
It's written by an "independant reviewer" because Newsforge didn't trust anyone on staff to qualify as unbiased.
He says nicer things about the product than I would
My main quibble with him is that he didn't factor in their history of suing their clients, but that's actually reasonably fair, as SCOX has so far only gone after deep pockets. (Still, I would consider it sufficient reason in and of itself to avoid the company.)
I think we've pushed this "anyone can grow up to be president" thing too far.
Probably the only thing you get is compatibility with older versions of UnixWare. Thus, there will be few if any new converts, but there might be some upgrades. Overall, if I were running UnixWare, I'd be looking at upgrading to some form of Linux instead.
Unknown host pong.
Out of all of them, SCO has always been the biggest nightmare to setup and sysadmin out of all of them.
- The SCO documentation is rubbish. It was spread over a huge number of volumes that took you hours to try and find the answer to any problem
- Bearing in mind that SCO's an x86 UNIX, the driver support is minimal
- No publishers have ever taken much interest in writing specific books for it. Aside from generic UNIX books, there's not a lot else compared to the very good books on all the UNIXes
- Even Evi Nemeth's "UNIX System Administration Handbook" (the UNIX bible for those who don't know) has never even mentioned it (at least in the 2nd & 3rd editions I have) whereas even IRIX and DEC OSF/1 get their own sections!
- I don't even remember it coming with a C compiler by default
IMHO SCO is UNIX from the Dark Ages.
Gentoo Linux - another day, another USE flag.
Looking at those pricing numbers, and the [lack of significant] advantage UnixWare has over, say, everything else, it seems that SCO is still stuck back in the glory days of Unix
They're not just stuck, they are dancing around a cauldron at night during a full moon, with a pack of naked lawyers wearing goat's heads trying to conjure back up the glory days.
"Eve of Destruction", it's not just for old hippies anymore...
You're watching too much FoxNews. "Objective" doesn't mean "slavish devotion to the subject". It means "without self-selecting bias regarding the subject". The human capacity for the true "objectivity" usually attributed to journalists and scientists is possible only perhaps in machines, or more likely in objects untouched by humans. But who wants to hear what they think?
--
make install -not war
Unix, one of the peaks of engineering history, has fallen into the grubby paws of this band of hapless Utah lawyers. Of course coding is deprioritized - that only costs money, and is extraneous to the SCO business model. Help us, Obi-Wan, you're our only hope!
--
make install -not war
and let's just put it this way, linux is MUCH MUCH MUCH MUCH better, hell, I'd use windows over that piece of shit. it's slow,(takes 10 minutes to boot on a 200 mhz) it's not flexible, lack of apps, and if you dont have the administrative password, you cant retrieve it. it also uses a ton of outdated shit as well, the version I played with still had a standard unix shell... argh I cant even begin to point out how fucked up it is.
it's a waste, and SCO knows this. this is why they want linux to be theirs, they get some stock, they get a top quality system they never made, and they want it to be exclusive to them. unixware is simply a hack of SYSV unix, and sco openserver is much the same way.
Hi. We're SCO. We don't believe in the GPL, but we include a host of GPL'd applications in our version of UNIX that no one other than those already using it (and those are just trying to move away from it) want.
How to get fired: recommend software from a vendor who's source is closed and may not be around in the near future. No... I don't mean Microsoft. I mean SCO.
The only thing necessary for Micro$oft to triumph is for a few good programmers to do nothing". North County Computers
If horseshit was seen growing legs by an evolutionist, they'd name the new species Darl.
:P
-5 Flamebait, but you can't hurt my karma
The only thing necessary for Micro$oft to triumph is for a few good programmers to do nothing". North County Computers
It's a huge pain in the ass. I've never seen a decent sized business (200+ people) without any software violations. It's just too hard to keep track of who owes what to whom and when it is going to expire. Not only is SCO's licensing expensive, it's pretty damn complicated too. Just look at the bottom of the article. The second half is all licensing details and I dare anyone to try and figure out their department's needs in less than an hour.
So yeah, it is expensive, but it also looks like a rat's nest.
Why I'm responding to a troll is beyond me, but I'll point out 2 reasons why even the trolls should RTFA.
According to SCO's own release and the review, a maximum of 8 processors are supported, not "scaling to hundreds of CPUs" as the parent states. Also, the review actually said more about SCO's products than I've ever gotten from SCO themselves, even back in '95 when I was looking for a UNIX for Intel (I chose Linux mainly because I couldn't find enough info on SCO, and the BSD documentation was something I wasn't able to make sense out of at the time). Admin GUIs are not something I expected from SCO, but apparently they're there. Their clustering technology is intriguing, and is another thing I didn't know they were even capable of.
If for no other reason than to "know your enemy" a good "technical" review of their product speaks more than any press on either sides of the lawsuits can for the company in the long run.
For those that must know, I run a number of servers, mostly Red hat ES 3.0 servers (including a 3 tier LVS cluster), with some Win 2k/2003 mixed in, and am writing this from a Powerbook running OS X. It's glad to know that is doesn't sound like SCO has made any jumps that would make me consider their product for work, so I need not fear the dark side.
I am, and always will be, an idiot. Karma: Coma (mostly effected by
The review seemed to be pretty vanilla, clued, thorough software review. The writer only looked only at the software that he was asked to review. Yeah, the review looked favorable -- as almost all software reviews do unless the software is total crap; reviewers tend to write about the good things. The part that I found really interesting was the level of detail (and the big numbers) included in the pricing information, most reviews show rather sketchy and incomplete pricing details. SCO prices everything ala-carte and seems to be going for the "we already got you by the balls" customers. What I took from the review was that the software is not necessarily crap but that you are going to pay for the privelege of using it.
There are a lot of interesting observations in the review, including:
We use Unixware 7 at work. It was the system recommended by a specific software vendor at the time (and that is the only application we run on the box).
/etc configure what services, this entire area of development is missing. Again, refer to points 1 and 2 and see what a nightmare this could potentially be.
We have been trying to identify the best migration plan for the following reasons:
- SCO's lack of hardware makes upgrading a nightmare of its own. With Windows and Linux, I can buy virtually anything (server hardware, that is) and expect it to "just work".
- The fact that SCO is at least at serious risk of collapse in the foreseeable future means that we now need to keep a copy of the hardware compatibility list and Unixware installation media in case of catastrophe (see point 1 and now imagine no tech support). This is a non-concern with any other reasonable alternative.
- Documentation sucks. From man pages either being non-existent or missing critical information such as what files in
- Related to the last point, Unixware expects you to use the scoadmin tool to do everything, including configure network cards. The location of even a basic ifconfig file is well hidden. To make this matter worse, scoadmin is non-intuitive to maneuver and also does not support termcap/terminfo -- you must use an ANSI terminal or the display will be garbled. Our vendor provides a custom telnet application to ensure you are always in ANSI.
- No support of PAM. We would like to simply integrate our logins with our Windows domain controller. Not possible with Unixware.
The very recent adoption of open source tools is actually the best thing they've done. In the version we have installed, SCO included VisionFS which provides SMB shares but is just not the same quality as Samba. More recent versions have dropped VisionFS and added more open source tools.
That's a quick review off the top of my head from somebody who uses it every day and looks forward to the day that we can be done with it.
"Host Bus Adapter, a kind of interface for SCSI systems."
A HBA can be SCSI, IDE or SATA.
But I doubt the SATA support is included in the base settlement; I mean user contract.
For add on packages or upgrades, just call the law offices of
Dewey, Screwam and Howe
I am the unwilling control for my Origin.
"Hmm, never heard of NeTraverse Merge... who develops it?"
Netraverse, of course. The Win4Lin people. Actually, Win4Lin and Merge are basically the same product.
"How does it compare with WINE?"
From a technical standpoint, we're talking apples and oranges. Wine is a project to independently implement a runtime environment that will be binary-compatible with Microsoft Windows. Win4Lin is an i386 virtualization tool tailored to run Microsoft Windows in a VM (virtual machine) on i386-based *nix.
From a practical standpoint, both are useful. Wine is, of course, free, while Win4Lin is a commercial product. Wine does not require any Microsoft software; Win4Lin requires you to provide MS Windows (to install and run in the VM). Wine is trying to chase Microsoft's moving target; Win4Lin lets you run the real thing. Wine uses less resources. Win4Lin is far more compatible -- it works with most any non-multimedia application flawlessly.
I use both. Win4Lin is extremely useful; it lets me run "the real thing" in a VM ("Windows in a window"), but with significantly better performance then VMware (doubtless because Win4Lin is tuned to just run Windows, while VMware is a full-blown, general-purpose VM). Wine yields better performance for applications which work with Wine. Win4Lin means no Wine compatability headaches; just install and run like a "real" 'doze box.
FWIW, IMO, YMMV, HTH, HAND, etc.
Here's the history behind Win4Lin/Netraverse, from my files:
It appears the company which originally developed the Merge software was "Locus Computing Corporation". They marketed a product called "DOS/Merge", which is the ancestor to the Win4Lin that we all know and love. DOS/Merge was later called "386/Merge" when 386 protected mode support was added.
At some point, a company called "Platinum" bought Locus. They apparently integrated Merge with other components into product lines called "PC-Enterprise" and "PC-Interface".
The Merge product was licensed to several other companies, including SCO, Sun, and HP. Sun and SCO both have commercial Unix products that run on Intel hardware; they offer "SCO Merge" and "Sun Merge" as layered products for their Unixes. (SCO, of course, later sold major assets (including their name) to Caldera, and Caldera then changed their name to SCO.)
At some point, a company called "DASCOM" bought the rights to Merge from Platinum. (Shortly thereafter, Platinum was bought by Computer Associates (CA), and fell off the Earth.) DASCOM was later bought by IBM. IBM was not interested in Merge, and spun the Merge group off as "TreLOS". TreLOS later merged with Lastfoot.com, and became "NeTraverse".
So:
Locus -> Platinum -> DASCOM -> IBM -> TreLOS + Lastfoot -> NeTraverse
DOS/Merge -> 386/Merge -> PC-Enterprise & PC-Interface -> Win4Lin
dragonhawk@iname.microsoft.com
I do not like Microsoft. Remove them from my email address.
What made strong impressions on me was events line:
client has 10 xenix boxes. Client gets some network cards because they FINALLY want them to talk to each other. So they spent several hundred $$$/machine for cards, a bunch of Coax and 10 copies of Xenix TCP/IP software. I got to install.
I spend the day working on the boxes, I'd pull the software, install, do the licensing, leave the license card. A little waiting for machines, so I run wires, and it's getting done.
Oh, but the machines (all?) spew an "alert" that there is a duplicate key in use.
Somewhere, I put the same key in twice.
We call SCO. We get told (on Mon) that someone will "call you back before Thursday."
Uh... no. surrounded by shrink wrap and a someone upset client...
No love. I have to uninstall everything, reinstall. Another several hours.
Next day, things network! Woo hoo! but...
They login by project name. But they can't RCP. or rlogin. We put on passwords (isolated network in a secure room, no passwords). Kinda a PITA.
Oh, project "pacific1" won't rsh/rlogin still. Nor a couple others. Still waiting for the "brand new customer" + VAR support call back.
The CAD support people come through (ArrisCAD rules!). Seems "8 letter login names won't work. We know, it's stupid; we agree. Oh, and you can't extract the TCP software license key," so if I keep waiting for support, they'll tell me to do what I did.
-----
This sort of action was repeated over and over. When, later, UnixWare (1992ish) was foisten on me, the hole bad hack of YP and mounting NFS and every painful step just burned into my brain more and more that this was a Unix half owned by Microsoft and its sole purpose was to make people like DOS and Windows 3.0
As soon as BSDI could run SCO binaries, I called the remaining (former clients) who still were stuck with SCO for some software lockin.
I will maintain that the ONLY reason SCO classic sold stuff through the late 80s was because of software that only ran on it. And those people got locked in because it was the only unix that could run on a 286 back in the day.
Move forward and the way to make money from SCO is to "pump and dump" - lawsuits about non-existent intellectual IP and the price goes up enough to sell a bunch of stock and pocket some cabbage.
Sure, the JFS that IBM brought from OS/2 came from SCO. Right, I'll get on that. And the rest of the rot.
Bad company that became obsolete (not EVERYBODY stopped innovating, mr sco) and got bought by a genius from Novell (remember when Novell I ruled the world doing the equiv of a stateful NFS and lpr for $10,000).
Evil company; costly yet mediocre software.
The clueless, pointless SCO baiting in this thread is depressing. You all hate them, but I don't think many of you know why. You gobble up propaganda and follow the flock like children. I'm so sick of uninformed narrow mindedness and sad fanboys that I rarely look at the discussions following Linux stories any more. How many ways are there to say "linux r0x0rs SCO is teh SUX!!!!". Don't you think we've got the message by now?
Yes, this is flamebait, but there are plenty of people round here who deserve flaming. Say something interesting or say nothing at all. We know the party line. We don't need you to trot it out again and underline it with a Monty Python quote.