And lots of lawyers will give you an hour or so of free advice/counseling to help you get started.
I know because I had to counter-sue my landlord once, and out of seven lawyers I spoke to in regard to the issue only one charged me. They mostly tell you things like, "You're on the right track, you may want to try this, or you have to file that at this time and with this dept." They'll also tell you if you've got a fair judge or not, and if they think you're doing a good job on your own, they'll tell you that too, which is kind of encouraging.
Much of the legal field isn't too hard to grok, if you want to handle it yourself, you just want to get advice from lawyers on what the filing processes are because a lot of the judges are too busy to answer your proceural questions in court.
Actually, there have got to be some legal firms already using Linux and other open source software. Maybe someone well known in the community could appeal to those firms for help.
Actually, Novell only said SCO didn't have rights to terminate IBM's license. Granted, it's a fine distinction and leads to the conclusion that SCO lacks standing, but it doesn't tell us with certaintuy whether IBM's current license is primarily with Novell or SCO.
If Novell *is* the primary owner of IBM's license, then SCO likely lacks standing to bring any case at all. I think we'll have to wait for Novell to say right out "We're the primary licensor for the IBM contract, not SCO," before that can play out, though.
Darl gets a big fat payoff if he can deliver four straight profitable quarters. Most of it is in stock, which means he'll have to keep up the fiasco for another quarter or two to cash out.
At that point, I think we can expect him to leave SCO. If there is any SCO left to leave. Maybe the final legal showdown will be Darl v. Ralph, to be filed in late 2004 or early 2005. We all know how much Darl loves to sue his employers.
Anyway, this means the SCO v. IBM case is not likely to ever make it to court because there's *no* motivation for Darl to go that far.
In the meantime, he'll do whatever it takes to show profit on the next two or three 10-Q's. He'll slash personnel, support, anything, doesn't matter how it affects SCO's long term prospects, as long as he shows profits each quarter. He'll try to get people to pay for SCO IP in Linux licenses NOW, not after the case is resolved in court, because he doesn't care what happens that far down the line.
He needs the money on the books and in the 10-Q next quarter and the following one. He's got two profitable quarters in a row, though he probably wouldn't have made it this quarter without cutting personnel and associated costs. Two more to go, and he's golden.
If he hasn't done it already, we can expect some *extremely* creative accounting over the next two quarters. Or more money from MS. MS, according to the latest 10-Q (available at SEC), has apparently purchased those "expanded licensing options" that were mentioned in the April 30 10-Q.
Darl's biggest fear is that something will shut down SCO and/or it's FUD machine within those next two quarters. If he sounds irrational and afraid, well, that's because he is. He can't pull any more profits out of Germany. Australia, Austria, and Poland are lining up to gag him in their countries. Red Hat's trying to do the same in the U.S. Of course, none of this matters much as long as no court decisions are reached within the next 3 quarters. Which means delay, delay, and delay will be SCO's legal strategy going forward.
That 95% figure is only for outstanding contracts that Novell had prior to selling its Unix businees to SCO. I would guess that the majority of SCO Unix licenseeing contracts require a far smaller royalty to Novell at this point.
On the other hand, the IBM license may be one of those contracts, in which case SCO might not even have the *standing* to sue IBM. I wonder if IBM is pursuing this angle. Some of the averments they deny in their response to SCO's amended complaint certainly leave them open to make this argument.
"So it's ok to threaten a community, but not ok for a member of that same community to stand up?"
Actually, that's probably true, but also probably irrelevant in this case. Red Hat's initial filing included enough press quotes from SCO management specifically mentioning Red Hat, that I think Red Hat's controversy claim will hold up.
SCO is saying hear that we never attacked Red Hat in the press, just Linux. But the record shows otherwise, lots of references to Red Hat by Darl & Co., so I suspect SCO will simply come off looking like hypocrites to the judge.
I hope so, anyway. I really, really, want this Red Hat suit to succeed and shut down SCO's FUD machine.
Red Hat's claim of there being an "actual controversy" should also hold up because of SCO's statements that they intend to invoice corporate users of Linux. Clearly some, probably most, of those corporate users are using Red Hat.
Unless it's 2 dozen OpenGL screensavers running unthrottled on the server.
Which reminds me, one important bit of advice for setting up a terminal server: make sure its got a good graphics card. Screen graphics (especially for MS and Citrix) are rendered on the terminal server before being forwarded to the client, where they are re-rendered for display. It seems like a no-brainer, but you'd be surprised at how many companies put terminal services on 3Ghz box with 2 GB RAM and 16 MB onboard video RAM, then wonder why the terminal server starts slowing down at 8 users. D'uh.
"They are still too expensive because the manufacturers get bulk discounts. There is no need for thin clients to cost more than Pentium 4 desktop PCs with HDD+DVD+Floppy drives, gobs of RAM and a 3D accelerator."
Christ, tell me where you're getting your PC's. As far as I know, a PC configured as you've described usually costs about 2-3 grand. I can price out a system pretty close to that configuration (with some marginal compromises) down to about $1000, but that's still 65% more expensive than the thin client. And all the client gets for that is functionality they don't need, functionality the users will use to waste time screwing around on the job, and higher support costs.
I'm doing it with Wyse 5440XL's and Dell 19" LCD's. Each part is a little less than $650, so the cost for each workstation is about $1250.
Sure, we could have done it for less, but the company owner likes the new LCD's and she decided that she was getting her savings on support costs, so a one-time initial investment to get good video, plus extra RAM and Linux support, was a marginal extra cost she was willing to spring for.
IOW, and this is directed at the other readers -- not the parent poster, who clearly gets this point already, she chose thin clients not because the hardware costs were cheaper, but so she could get more of what she needs at the same price, and lower support costs going forward.
Actually, the primary reason for moving to decentralized was because it was cheaper and more robust.
That equation has now changed. It's cheaper to run a centralized model once again. And robustness has increased in the centralized model, but decreased in the decentralized model due to internet access, in particular "all the cool things" end users now want to download and install on their PC's.
People can talk about securing the workstations better, but the truth is that it's just easier to secure thin clients.
This is coming from someone who has to support both models for different clients. I always recommend thin clients now for companies just starting up. Small companies that already have an infrastructure in place want to get an ROI on their investment first, so you have to find ways make their PC infrastructure more flexible, then slowly introduce the technology.
"I want to know why the X capable clients cost so much more than the Winterm clients."
They don't. The HP t5700 starts at $599. A similarly configured Wyse 5540XL costs about $629, runs SuSE Linux 7.x, which makes it X compatible, and also has RDP and ICA clients.
In other words, it's much more flexible than the HP unit, costs about the same, and runs X.
Because they're talking about the t5700 which starts at $599, comes with 192 MB RAM, ATI Rage XC 8MB RAM, 4 USB ports, and embedded XP.
You're right though. The article itself references the 733Mhz processor, which is the core of the t5500, starts at $369, and comes with: 32MB RAM, ATI Rage XC 8MB RAM, 4 USB ports, and WinCE.
I personally wouldn't recommend either, because given the way the market is going it would be stupid to lock in on the MS platform rather than getting a system that can run both X and MS Terminal Services.
However, if someone *is* determined to stick with Microsoft and doesn't want the option to upgrade to LTSP, then the t5500 is still going to be slow, due as much to its low RAM as the processor, and nowhere near as flexible as the t5700.
"Even Windows these days has "terminal services" that works (or at least thats the rumor) in a similar fasion."
It does work remarkably well. Especially in a small-midsized office environment. But it's expensive, not prohibitively, but more so than it should be. And for a large corporate environment, once you get over 150-200 users, you're probably better off adding Citrix to the mix, which adds even more costs.
Still, most companies are going to insist on MS, in part due to compatibility with their partners, in part because it's still easier and cheaper to find technical support personel or consultants that are familiar with MS as opposed to Linux. That's changing, but not quickly.
True, you can set up a 486 with network boot and have a functional XTerminal.
However, and it's not necessarily as stupid as it sounds, some companies upgrade still functional equipment because the old equipment looks like shit after 3 or 4 years of getting beat up and collecting dust. It's like replacing a threadbare office chair.
And in that case a new TC is going to look a lot better than an ancient 486, P1, or P2 behomoth. And it will be cheaper than a *good* new PC.
As I said earlier in the thread, I recently spent about $650 a piece for some TC's. But they had more RAM than most TC's and were running Linux instead of WinCE or embedded XP. They also had sound, USB support, and significantly better graphics capability than the cheaper TC's. Even with Linux being free, it's still worth paying more to get it, since it gives me the flexibility to add LTSP to the mix later on.
Also, I imagine these units from HP will be running embedded XP, which probably pushes the price up significantly. ViaC3 clients with 32 MB RAM running WinCE are going to be much cheaper, but they are also not good for anything but running RDP sessions, and they usually have crappy graphics (8 bit color at 1024x768 resolution). The higher price for the HP unit probably reflects cost increases for more RAM, better graphics, sound, USB support, and embedded XP.
"The decrease in administration hassles, the improved security, the decreased power consumption and the interchangeability all add up to a pretty compelling advantage compared to putting a full-blown PC on every desk - for a medium to large organization."
Not to mention the dereased costs for hardware support. No need to worry about the client screwing up their hard drive for instance, or having it fail.
And it's not just medium-large organizations that can save money on this kind of setup. Any company that's just starting up can save a lot of money by deploying a thin client solution with LTSP or even MS Terminal Server. The initial costs may be about the same, but the savings going forward from that point are quite significant. I had one small client (about 5 users) with this setup that at one point went nearly a year w/o needing support.
Actually, the fast ones usually run P3's. I expect that will change to Pentium M class processors as the price for them comes down. Those really are nice chips, and usually get the same processing speed at 1/2 to 2/3 the clock rate of a similarly quick P4.
Um, on a thin client, compilation and compression are going to take place on the server, not the client. So having a faster processor on the client isn't going to help you at all.
Well, you can get a Thin Client for as little as $200-$400. I just bought several for one of my clients at about $650 a pop, but that was because I wanted them to be capable of Windows emulation and X Windows.
The real savings is in support costs, though. With regualar PC's and hard drives, the initial costs to setup and secure the workstation are much higher, and even then the users are always screwing things up with Bonzia Buddy, assorted screensavers, etc. Using thin clients with Linux or Terminal Server really cuts down on support costs.
Anyway, the statement in that article which I found odd was that HP was the leader in the Thin Client market. Everytime I have to set clients up with more TC's, I research the market again, and Wyse is always the best deal. Frankly, I thought they were in the lead for the thin client market, though I may be wrong.
"...the details of Sequent's contract are not known publicly, yet"
Well, they are actually, as referenced above. At the least they are known as far as the courts know them and these are the documents SCO intends to rest its case on.
They are basically identical to the first two licenses that IBM signed with ATT, Exhibits A & B, at the same site. Sequent did not have an additional license such as IBM's sideletter agreement, but as part of a clarification request from Berkely, ATT sent out letters to all the System V licensees, in the mid-late 80's I believe, granting them all "ownership of derivative works created by or for them". Essentially the same wording as section 2 of IBM's sideletter agreement.
Sequent signed it's license with ATT in '85 and would have been one of the recipients of this letter, which SCO interestingly enough, has not included amongst its submitted evidentiary exhibits.
So, Sequent had the same rights to its own derivative work as IBM, despite not having a sideletter agreement.
"Until you know what that contract says, you would be foolish to dismiss this issue as moot."
Well, seeing as SCO has submitted two contracts with Sequent to the courts, and that I've read them, I think that the conditions you've set for me being foolish to regard the issue as moot have been over-ridden. You may argue that we don't know yet whether SCO has other contracts to refute these points, and my response would be that neither do the courts, since they evidently haven't been submitted as exhibits.
"So are you saying that rights Sequent did not have were transferred to IBM when Sequent was purchased?"
Ah. Now I get it: you're a troll. That's a really twisted way of interpreting my argument. What I am saying is that even if Sequent hadn't already rec'd those rights from ATT, Sequent would have inherited the IBM's rights when it was incorporated int IBM's corporate structure.
Please don't bother us again with you're "insights" until you've actually read the contracts.
Slashdot Mirror
Well, it cracked me up. But then I know most of the basic rules of grammar, punctuation, and spelling.
Maybe you're just insecure: is it 'cause ya ain't got over that 370 on the verbal part of the SAT yet?
And lots of lawyers will give you an hour or so of free advice/counseling to help you get started.
I know because I had to counter-sue my landlord once, and out of seven lawyers I spoke to in regard to the issue only one charged me. They mostly tell you things like, "You're on the right track, you may want to try this, or you have to file that at this time and with this dept." They'll also tell you if you've got a fair judge or not, and if they think you're doing a good job on your own, they'll tell you that too, which is kind of encouraging.
Much of the legal field isn't too hard to grok, if you want to handle it yourself, you just want to get advice from lawyers on what the filing processes are because a lot of the judges are too busy to answer your proceural questions in court.
Actually, there have got to be some legal firms already using Linux and other open source software. Maybe someone well known in the community could appeal to those firms for help.
Actually, Novell only said SCO didn't have rights to terminate IBM's license. Granted, it's a fine distinction and leads to the conclusion that SCO lacks standing, but it doesn't tell us with certaintuy whether IBM's current license is primarily with Novell or SCO.
If Novell *is* the primary owner of IBM's license, then SCO likely lacks standing to bring any case at all. I think we'll have to wait for Novell to say right out "We're the primary licensor for the IBM contract, not SCO," before that can play out, though.
Darl gets a big fat payoff if he can deliver four straight profitable quarters. Most of it is in stock, which means he'll have to keep up the fiasco for another quarter or two to cash out.
At that point, I think we can expect him to leave SCO. If there is any SCO left to leave. Maybe the final legal showdown will be Darl v. Ralph, to be filed in late 2004 or early 2005. We all know how much Darl loves to sue his employers.
Anyway, this means the SCO v. IBM case is not likely to ever make it to court because there's *no* motivation for Darl to go that far.
In the meantime, he'll do whatever it takes to show profit on the next two or three 10-Q's. He'll slash personnel, support, anything, doesn't matter how it affects SCO's long term prospects, as long as he shows profits each quarter. He'll try to get people to pay for SCO IP in Linux licenses NOW, not after the case is resolved in court, because he doesn't care what happens that far down the line.
He needs the money on the books and in the 10-Q next quarter and the following one. He's got two profitable quarters in a row, though he probably wouldn't have made it this quarter without cutting personnel and associated costs. Two more to go, and he's golden.
If he hasn't done it already, we can expect some *extremely* creative accounting over the next two quarters. Or more money from MS. MS, according to the latest 10-Q (available at SEC), has apparently purchased those "expanded licensing options" that were mentioned in the April 30 10-Q.
Darl's biggest fear is that something will shut down SCO and/or it's FUD machine within those next two quarters. If he sounds irrational and afraid, well, that's because he is. He can't pull any more profits out of Germany. Australia, Austria, and Poland are lining up to gag him in their countries. Red Hat's trying to do the same in the U.S. Of course, none of this matters much as long as no court decisions are reached within the next 3 quarters. Which means delay, delay, and delay will be SCO's legal strategy going forward.
That 95% figure is only for outstanding contracts that Novell had prior to selling its Unix businees to SCO. I would guess that the majority of SCO Unix licenseeing contracts require a far smaller royalty to Novell at this point.
On the other hand, the IBM license may be one of those contracts, in which case SCO might not even have the *standing* to sue IBM. I wonder if IBM is pursuing this angle. Some of the averments they deny in their response to SCO's amended complaint certainly leave them open to make this argument.
"So it's ok to threaten a community, but not ok for a member of that same community to stand up?"
Actually, that's probably true, but also probably irrelevant in this case. Red Hat's initial filing included enough press quotes from SCO management specifically mentioning Red Hat, that I think Red Hat's controversy claim will hold up.
SCO is saying hear that we never attacked Red Hat in the press, just Linux. But the record shows otherwise, lots of references to Red Hat by Darl & Co., so I suspect SCO will simply come off looking like hypocrites to the judge.
I hope so, anyway. I really, really, want this Red Hat suit to succeed and shut down SCO's FUD machine.
Red Hat's claim of there being an "actual controversy" should also hold up because of SCO's statements that they intend to invoice corporate users of Linux. Clearly some, probably most, of those corporate users are using Red Hat.
Wyse, and I got it wrong. They use AMD processors. My bad. So it's not Intel, but it's not Via C3 either.
Unless it's 2 dozen OpenGL screensavers running unthrottled on the server.
Which reminds me, one important bit of advice for setting up a terminal server: make sure its got a good graphics card. Screen graphics (especially for MS and Citrix) are rendered on the terminal server before being forwarded to the client, where they are re-rendered for display. It seems like a no-brainer, but you'd be surprised at how many companies put terminal services on 3Ghz box with 2 GB RAM and 16 MB onboard video RAM, then wonder why the terminal server starts slowing down at 8 users. D'uh.
It's just so courageous to be a smart ass when you're an Anonymous Coward.
Dickless Asshole.
"They are still too expensive because the manufacturers get bulk discounts. There is no need for thin clients to cost more than Pentium 4 desktop PCs with HDD+DVD+Floppy drives, gobs of RAM and a 3D accelerator."
Christ, tell me where you're getting your PC's. As far as I know, a PC configured as you've described usually costs about 2-3 grand. I can price out a system pretty close to that configuration (with some marginal compromises) down to about $1000, but that's still 65% more expensive than the thin client. And all the client gets for that is functionality they don't need, functionality the users will use to waste time screwing around on the job, and higher support costs.
I'm doing it with Wyse 5440XL's and Dell 19" LCD's. Each part is a little less than $650, so the cost for each workstation is about $1250.
Sure, we could have done it for less, but the company owner likes the new LCD's and she decided that she was getting her savings on support costs, so a one-time initial investment to get good video, plus extra RAM and Linux support, was a marginal extra cost she was willing to spring for.
IOW, and this is directed at the other readers -- not the parent poster, who clearly gets this point already, she chose thin clients not because the hardware costs were cheaper, but so she could get more of what she needs at the same price, and lower support costs going forward.
Actually, the primary reason for moving to decentralized was because it was cheaper and more robust.
That equation has now changed. It's cheaper to run a centralized model once again. And robustness has increased in the centralized model, but decreased in the decentralized model due to internet access, in particular "all the cool things" end users now want to download and install on their PC's.
People can talk about securing the workstations better, but the truth is that it's just easier to secure thin clients.
This is coming from someone who has to support both models for different clients. I always recommend thin clients now for companies just starting up. Small companies that already have an infrastructure in place want to get an ROI on their investment first, so you have to find ways make their PC infrastructure more flexible, then slowly introduce the technology.
4) Profit!
I know "me too" posts are frowned upon, but i just have to say:
Exactly.
Just thought you'd like to know that there are others here who agree with you.
"I want to know why the X capable clients cost so much more than the Winterm clients."
They don't. The HP t5700 starts at $599. A similarly configured Wyse 5540XL costs about $629, runs SuSE Linux 7.x, which makes it X compatible, and also has RDP and ICA clients.
In other words, it's much more flexible than the HP unit, costs about the same, and runs X.
Because they're talking about the t5700 which starts at $599, comes with 192 MB RAM, ATI Rage XC 8MB RAM, 4 USB ports, and embedded XP.
You're right though. The article itself references the 733Mhz processor, which is the core of the t5500, starts at $369, and comes with: 32MB RAM, ATI Rage XC 8MB RAM, 4 USB ports, and WinCE.
I personally wouldn't recommend either, because given the way the market is going it would be stupid to lock in on the MS platform rather than getting a system that can run both X and MS Terminal Services.
However, if someone *is* determined to stick with Microsoft and doesn't want the option to upgrade to LTSP, then the t5500 is still going to be slow, due as much to its low RAM as the processor, and nowhere near as flexible as the t5700.
"Even Windows these days has "terminal services" that works (or at least thats the rumor) in a similar fasion."
It does work remarkably well. Especially in a small-midsized office environment. But it's expensive, not prohibitively, but more so than it should be. And for a large corporate environment, once you get over 150-200 users, you're probably better off adding Citrix to the mix, which adds even more costs.
Still, most companies are going to insist on MS, in part due to compatibility with their partners, in part because it's still easier and cheaper to find technical support personel or consultants that are familiar with MS as opposed to Linux. That's changing, but not quickly.
True, you can set up a 486 with network boot and have a functional XTerminal.
However, and it's not necessarily as stupid as it sounds, some companies upgrade still functional equipment because the old equipment looks like shit after 3 or 4 years of getting beat up and collecting dust. It's like replacing a threadbare office chair.
And in that case a new TC is going to look a lot better than an ancient 486, P1, or P2 behomoth. And it will be cheaper than a *good* new PC.
As I said earlier in the thread, I recently spent about $650 a piece for some TC's. But they had more RAM than most TC's and were running Linux instead of WinCE or embedded XP. They also had sound, USB support, and significantly better graphics capability than the cheaper TC's. Even with Linux being free, it's still worth paying more to get it, since it gives me the flexibility to add LTSP to the mix later on.
Also, I imagine these units from HP will be running embedded XP, which probably pushes the price up significantly. ViaC3 clients with 32 MB RAM running WinCE are going to be much cheaper, but they are also not good for anything but running RDP sessions, and they usually have crappy graphics (8 bit color at 1024x768 resolution). The higher price for the HP unit probably reflects cost increases for more RAM, better graphics, sound, USB support, and embedded XP.
"The decrease in administration hassles, the improved security, the decreased power consumption and the interchangeability all add up to a pretty compelling advantage compared to putting a full-blown PC on every desk - for a medium to large organization."
Not to mention the dereased costs for hardware support. No need to worry about the client screwing up their hard drive for instance, or having it fail.
And it's not just medium-large organizations that can save money on this kind of setup. Any company that's just starting up can save a lot of money by deploying a thin client solution with LTSP or even MS Terminal Server. The initial costs may be about the same, but the savings going forward from that point are quite significant. I had one small client (about 5 users) with this setup that at one point went nearly a year w/o needing support.
Actually, the fast ones usually run P3's. I expect that will change to Pentium M class processors as the price for them comes down. Those really are nice chips, and usually get the same processing speed at 1/2 to 2/3 the clock rate of a similarly quick P4.
Um, on a thin client, compilation and compression are going to take place on the server, not the client. So having a faster processor on the client isn't going to help you at all.
Well, you can get a Thin Client for as little as $200-$400. I just bought several for one of my clients at about $650 a pop, but that was because I wanted them to be capable of Windows emulation and X Windows.
The real savings is in support costs, though. With regualar PC's and hard drives, the initial costs to setup and secure the workstation are much higher, and even then the users are always screwing things up with Bonzia Buddy, assorted screensavers, etc. Using thin clients with Linux or Terminal Server really cuts down on support costs.
Anyway, the statement in that article which I found odd was that HP was the leader in the Thin Client market. Everytime I have to set clients up with more TC's, I research the market again, and Wyse is always the best deal. Frankly, I thought they were in the lead for the thin client market, though I may be wrong.
Phreidom,
My apologies, I thought I was discussing the issue with someone who had already done the research.
The Sequent contracts are available at SCO's website as Exhibits F & G. the URLS are
http://www.sco.com/ibmlawsuit/exhibitf.pdf
and
http://www.sco.com/ibmlawsuit/exhibitg.pdf
"...the details of Sequent's contract are not known publicly, yet"
Well, they are actually, as referenced above. At the least they are known as far as the courts know them and these are the documents SCO intends to rest its case on.
They are basically identical to the first two licenses that IBM signed with ATT, Exhibits A & B, at the same site. Sequent did not have an additional license such as IBM's sideletter agreement, but as part of a clarification request from Berkely, ATT sent out letters to all the System V licensees, in the mid-late 80's I believe, granting them all "ownership of derivative works created by or for them". Essentially the same wording as section 2 of IBM's sideletter agreement.
Sequent signed it's license with ATT in '85 and would have been one of the recipients of this letter, which SCO interestingly enough, has not included amongst its submitted evidentiary exhibits.
So, Sequent had the same rights to its own derivative work as IBM, despite not having a sideletter agreement.
"Until you know what that contract says, you would be foolish to dismiss this issue as moot."
Well, seeing as SCO has submitted two contracts with Sequent to the courts, and that I've read them, I think that the conditions you've set for me being foolish to regard the issue as moot have been over-ridden. You may argue that we don't know yet whether SCO has other contracts to refute these points, and my response would be that neither do the courts, since they evidently haven't been submitted as exhibits.
"So are you saying that rights Sequent did not have were transferred to IBM when Sequent was purchased?"
Ah. Now I get it: you're a troll. That's a really twisted way of interpreting my argument. What I am saying is that even if Sequent hadn't already rec'd those rights from ATT, Sequent would have inherited the IBM's rights when it was incorporated int IBM's corporate structure.
Please don't bother us again with you're "insights" until you've actually read the contracts.
Elmer's Glue and Borax.
Slow news day, huh?