Well, no, because your right to "use" the code also comes from the GPL. So if you give up the GPL, any internal "use" throws you into the whole IP "piracy" discussion.
Then you have to start rangling with your lawyers about the meaning of things in quotes.
I'm not sure that this is an accurate depiction. Reading the letter gives me no indication that SCO has had any hand in this code review. SGI has the complete SysV code and apparently performed the review purely as a result of public statements.
Thus, I am doubtfull that SCO has learned anything yet.
Yeah, I agree with those problems, and they need to be seen to. However, without the code, or at least something claimed to be the code, we can't start on the process. [We can however start on the process of setting up the process.]
What I would predict you'll find, if you were to ask, Is that some (maybe most) of the states/counties/precincts using these things don't have the code and can't get it. There is a basic flaw in the relationship. I suggest that sunshine laws can be used to pry at that relationship until we can do something about it.
In one of the earlier Diebold articles they found that the system was keeping double books. If you asked for the state total you got a report out of database #2, but if you asked for any one county you got a report out of the original database. That way if you did a spot check on any county, you would match the paper trail. However, nothing checked to be sure that adding up all the counties gave you the state total. The "state total" was considered the final result.
I wonder if we can argue before the court that the recall should not be delayed, because the electronic voting machines aren't any better than the punch cards.
I don't care when or if the recall takes place. What I want is for the courts to demand a better standard of quality in the election process.
While I do believe that Paper ballots and optical scanners are the correct solution, it does need to be pointed out that they suffer from many of the same flaws. The same companies make the optical scanner databases as the touchscreen ones, and they use the same databases.
The votes could still be manipulated during the count. You could still have the issue where it reports different numbers by county than for the whole state, hiding from a limited recount. Only a total manual recount would catch cheating.
The reason that paper is better as an immediate medium is that the original is human readable and verifiable. A printed reciept is only good if every voter carefully checks it. If it is large enough to be human usable you might as well print the whole thing.
Finally papaer and permanent markers are cheap and portable. California could run their recall election on paper and have the votes scanned in the next county. One day delay in counting is way less than six months to buy touch screens.
They priced the ads way too low. Heck, they shouldn't have "priced" them at all. They need to treat this as a product-placement and negotiate each ad in secret.
I know they need money fast, and that 9.2 is almost out the door. However, they are setting a bad precedent for themselves and pricing themselves into the "inferior goods" category.
Maybe $7000 is actually a reasonable price for something which is refreshed every few months. However, it looks cheep.
They are not selling newspaper ads or website banners and they need to stop acting like they are.
They need to fix this before masses of inferior ads drive their customers away.
I'm not real concerned about the actual ads. They'll be wiped by my custom settings anyway. However, since I use and like Mandrake, I'd prefer if they didn't screw their long term image in search of short term funds.
The question in my mind is, if the findings of the AT&T/BSD case are relevant--and I suspect they are-- does that mean they can be unsealed for this case?
I didn't mean to imply that we should be idle. I don't believe that the open source community is being idle. I think we're doing largely the right thing purely as a side effect of being a loose decentralized body. We put out the fires as we get to them. Each person extinguishes a small fire, or to go back to the cliff metaphor, gives a small push.
Information kills FUD. Quiet, calm sense goes a long way to making people listen.
What I've read seems to suggest that companies are watching the battle but most are not terribly concerned. Those that are concerned are worried about liability. If you can calm their liability fears it all goes back to the old arguments.
Businesses are concerned about liability because they think of software as a monolithic entity. They are not used to the conglomerate nature of OSS. Businesses think they are considering Linux when they really are not. They are considering Apache, or Samba, or KDE. They don't know where one begins and the next one ends. They only know the Linux hype.
The hype got them interested, not informed. A quiet word from a good friend can inform them. So maybe they end up on FreeBSD. They are out of the stream of Linux hype and out of the FUD. They can then experience the power of collaboration in a safe spot.
The original question is how aggressive should we be. The question is not should we be informed, both technically and legally. Yes, we should be. We should seek council. But that does not mean we should charge out blindly like an enraged bull.
IBM has lots of lawyers. Their lawyers say wait.
We can get lawyers too. Some already have. Judging by the quiet, my guess is that they also said wait.
The legal process specifically leaves time to gather facts. Our stroll to the courthouse does not give up that right. There's no rush.
Let's help a few old businesses cross the street while we're at it. Those we don't get to will still be there after SCO (unless they aren't).
That's right, it's called "evolution": more lazy will die, more aggressive will survive.
This is not actually true. High agression actually carries a large cost. In places without strong competition for a limited resource, high agression brings limited return for a high mortality rate.
While I could lecture for hours and swamp you with actual evolutionary examples (start with the sloth) I'll just let you think about High School. The high agression males who took up street racing and then (if they lived) joined the Marines vs the geeks who got food delivered. Compare their death-rates or insurance rates. [Ignore the eventual heart disease, they will have reproduced by then. You're not worth as much after you stop breeding. (aside: memes vs genes)]
Back on topic: Sitting back and ignoring them is a reasonable tactic. It doesn't take much effort. It doesn't take much money. We can spend our time wooing mates (like that sexy PPC970). All we have to do is wait for SCO to run off a cliff. If somehow they survive the fall, we drop rocks on them. Safe and sound on the higher ground.
Finally, OSS continues to gain in public exposure and sympathy (except MO state government). The later the trial takes place, the more likely a judge is to be favorably inclined toward us. Thus it is SCO's best interest to get this over as quickly as possible (not to mention legal fees.) It is our best interest to let public opinion and technical excellence build. SCO can bleed for a while.
Becoming a member actually was the route I chose. However, It occurs to me that purchasing a copy at a store actually has some benefit-- precisely because not all the money goes to Mandrake.
Stores stock what people buy. If Walmart and Staples see that a distro is selling, they wil put it on their shelves where it can be seen. If it sells enough, they will start thinking about selling Linux versions of third party software. Imagine what would happen if Walmart calls up Blizard and asks for the Linux version of something. That won't happen if the store doesn't make any money.
Okay; I give up. I don't see where it says that they can't do that.
The rules for consumers say they have to stop after three months.
http://www.ftc.gov/bcp/conline/pubs/tmarkg/donot ca ll.htm
The rules for business say they have to scrub at least every 90 days, but not which numbers. If they poll the database more often they get a period of time in which to call. If they poll everyday, they know exactly what day is the last day they can call you.
http://www.ftc.gov/bcp/conline/pubs/buspubs/call in g.htm
Now, I'm not suggesting they aren't violeting the intent of the rule, but they would not be violating the rules as described on the website.
Everyone is talking about the exemptions. No-one is talking about the delay.
If you sign up before August 1st, telemarketers have to stop calling you in October. This suggests that telemarketers should/will use the list as a "call now" list during September. Similarly, if you sign up after Aug 1, they have three months to call you as many times as posible.
Yes there has to be some delay, but three months?
I'm not saying they will all do this, but I bet some will.
As the husband of an ecologist [HOE], I must remind you that the order of living systems comes not from individuals working togeather, but from the less fit dying before they can reproduce.
The thing I forgot to hit, when I slipped into "speech" mode, is that your virus scan example differs from the RCU issue in an important way. In your example MS licenses AntiV from somebody. It's a contract about the AntiV. In the RCU case IBM licenses SysV from SCO/AT&T. RCU not only isn't in the contract, it didn't exist at the time of writing. RCU isn't "derived" from SysV in the traditional sense.
So the question is more akin to "If I invent a spoiler for a Honda, does that mean Honda can stop me from mounting it on a Yugo?"
As MS has shown, what is and isn't "part" of an OS is up for interpretation. [See IE vs Mozilla] Beyond that we have the question of "Which OS?"
[ignoring historical names] IBM used SysV to make AIX. IBM wrote RCU. IBM then added RCU to AIX [not SysV]. Is RCU a "part" of SysV?
It's development didn't require SysV. [It was done as a theoretical exercise. Followed by multiple implementations.]
It's execution doesn't depend on SysV. [It runs on/in Linux, AIX and OS/400. It doesn't run on or in the SysV code they liscenced from AT&T.]
RCU was never shipped with SysV [at least not at the point they wrote it]. They didn't add something to an existing operating system. They created a new OS based on an old one. Is Internet Explorer "part" of Windows 3.1? Win95? MacOS 9?
Imagine if you modified IE to take pluggins (yes it's already been done.) MS might be able to claim ownership of the "socket" but they can't claim ownership of everything that might get plugged into it. A license for derivatives of SysV would only cover the changes to SysV which were made so that it could acomodate the RCU code. Not the RCU code.
SCO seems to be claiming that this is wrong. That they own RCU, because RCU fits in that socket. No matter the origin of RCU. This is akin to XFree claiming ownership of KDE. Or KDE claiming ownership of Karamba. This would totally destroy everything we believe about the way interoperation and extension works. Adobe would own every photoshop pluggin.
Because this is a contract issue not a statute issue, I'm being extreme. However, the concept of ownership in these cases is not trivial. The concept of derivation is under attack. If you expand it the way SCO appears to be arguing you get absurdity. Therefor, it must fail. For the sake of our livelyhoods it must be shown to fail.
Slashdot Mirror
3rd variant:
Party B provided the good or service and Party A never paid. Party A has to give it back or pay some court negotiated restitution.
This is the situation SCO is in.
Well, no, because your right to "use" the code also comes from the GPL. So if you give up the GPL, any internal "use" throws you into the whole IP "piracy" discussion.
Then you have to start rangling with your lawyers about the meaning of things in quotes.
I work for a state government. We do it all the time. Unfortunately we tell them they have to use IE v6.0 on MSWindows with all the options turned on.
The server actually throws a runtime error if you try to use anything else.
Yep, gotta love bueaucracy.
Also remember that the moon is actually a very dark dark grey. It just looks bright because it is sitting in front of a "black" void.
I'm not sure that this is an accurate depiction. Reading the letter gives me no indication that SCO has had any hand in this code review. SGI has the complete SysV code and apparently performed the review purely as a result of public statements.
Thus, I am doubtfull that SCO has learned anything yet.
Yeah, I agree with those problems, and they need to be seen to. However, without the code, or at least something claimed to be the code, we can't start on the process. [We can however start on the process of setting up the process.]
What I would predict you'll find, if you were to ask, Is that some (maybe most) of the states/counties/precincts using these things don't have the code and can't get it. There is a basic flaw in the relationship. I suggest that sunshine laws can be used to pry at that relationship until we can do something about it.
In one of the earlier Diebold articles they found that the system was keeping double books. If you asked for the state total you got a report out of database #2, but if you asked for any one county you got a report out of the original database. That way if you did a spot check on any county, you would match the paper trail. However, nothing checked to be sure that adding up all the counties gave you the state total. The "state total" was considered the final result.
I wonder if we can argue before the court that the recall should not be delayed, because the electronic voting machines aren't any better than the punch cards.
I don't care when or if the recall takes place. What I want is for the courts to demand a better standard of quality in the election process.
Can't we use Freedom pf Information Act or sunshine laws to demand the code to the voting machines and then review them wether they like it or not?
While I do believe that Paper ballots and optical scanners are the correct solution, it does need to be pointed out that they suffer from many of the same flaws. The same companies make the optical scanner databases as the touchscreen ones, and they use the same databases.
The votes could still be manipulated during the count. You could still have the issue where it reports different numbers by county than for the whole state, hiding from a limited recount. Only a total manual recount would catch cheating.
The reason that paper is better as an immediate medium is that the original is human readable and verifiable. A printed reciept is only good if every voter carefully checks it. If it is large enough to be human usable you might as well print the whole thing.
Finally papaer and permanent markers are cheap and portable. California could run their recall election on paper and have the votes scanned in the next county. One day delay in counting is way less than six months to buy touch screens.
They priced the ads way too low. Heck, they shouldn't have "priced" them at all. They need to treat this as a product-placement and negotiate each ad in secret.
I know they need money fast, and that 9.2 is almost out the door. However, they are setting a bad precedent for themselves and pricing themselves into the "inferior goods" category.
Maybe $7000 is actually a reasonable price for something which is refreshed every few months. However, it looks cheep.
They are not selling newspaper ads or website banners and they need to stop acting like they are.
They need to fix this before masses of inferior ads drive their customers away.
I'm not real concerned about the actual ads. They'll be wiped by my custom settings anyway. However, since I use and like Mandrake, I'd prefer if they didn't screw their long term image in search of short term funds.
I think they already fired all of the people who do real work.
IBM has to have the code. They can do it as soon as their lawyers give the okay.
The penguin sometimes self distructs, leaving the finger only singed. Is that a win?
The question in my mind is, if the findings of the AT&T/BSD case are relevant--and I suspect they are-- does that mean they can be unsealed for this case?
Can we finally find out that bit of history?
I didn't mean to imply that we should be idle. I don't believe that the open source community is being idle. I think we're doing largely the right thing purely as a side effect of being a loose decentralized body. We put out the fires as we get to them. Each person extinguishes a small fire, or to go back to the cliff metaphor, gives a small push.
Information kills FUD. Quiet, calm sense goes a long way to making people listen.
What I've read seems to suggest that companies are watching the battle but most are not terribly concerned. Those that are concerned are worried about liability. If you can calm their liability fears it all goes back to the old arguments.
Businesses are concerned about liability because they think of software as a monolithic entity. They are not used to the conglomerate nature of OSS. Businesses think they are considering Linux when they really are not. They are considering Apache, or Samba, or KDE. They don't know where one begins and the next one ends. They only know the Linux hype.
The hype got them interested, not informed. A quiet word from a good friend can inform them. So maybe they end up on FreeBSD. They are out of the stream of Linux hype and out of the FUD. They can then experience the power of collaboration in a safe spot.
The original question is how aggressive should we be. The question is not should we be informed, both technically and legally. Yes, we should be. We should seek council. But that does not mean we should charge out blindly like an enraged bull.
IBM has lots of lawyers. Their lawyers say wait.
We can get lawyers too. Some already have. Judging by the quiet, my guess is that they also said wait.
The legal process specifically leaves time to gather facts. Our stroll to the courthouse does not give up that right. There's no rush.
Let's help a few old businesses cross the street while we're at it. Those we don't get to will still be there after SCO (unless they aren't).
ACs don't have foe lists
This is not actually true. High agression actually carries a large cost. In places without strong competition for a limited resource, high agression brings limited return for a high mortality rate.
While I could lecture for hours and swamp you with actual evolutionary examples (start with the sloth) I'll just let you think about High School. The high agression males who took up street racing and then (if they lived) joined the Marines vs the geeks who got food delivered. Compare their death-rates or insurance rates. [Ignore the eventual heart disease, they will have reproduced by then. You're not worth as much after you stop breeding. (aside: memes vs genes)]
Back on topic:
Sitting back and ignoring them is a reasonable tactic. It doesn't take much effort. It doesn't take much money. We can spend our time wooing mates (like that sexy PPC970). All we have to do is wait for SCO to run off a cliff. If somehow they survive the fall, we drop rocks on them. Safe and sound on the higher ground.
Finally, OSS continues to gain in public exposure and sympathy (except MO state government). The later the trial takes place, the more likely a judge is to be favorably inclined toward us. Thus it is SCO's best interest to get this over as quickly as possible (not to mention legal fees.) It is our best interest to let public opinion and technical excellence build. SCO can bleed for a while.
I'm really not in a rush.
Becoming a member actually was the route I chose. However, It occurs to me that purchasing a copy at a store actually has some benefit-- precisely because not all the money goes to Mandrake.
Stores stock what people buy. If Walmart and Staples see that a distro is selling, they wil put it on their shelves where it can be seen. If it sells enough, they will start thinking about selling Linux versions of third party software. Imagine what would happen if Walmart calls up Blizard and asks for the Linux version of something. That won't happen if the store doesn't make any money.
Okay; I give up. I don't see where it says that they can't do that.
The rules for consumers say they have to stop after three months.
http://www.ftc.gov/bcp/conline/pubs/tmarkg/dono
The rules for business say they have to scrub at least every 90 days, but not which numbers. If they poll the database more often they get a period of time in which to call. If they poll everyday, they know exactly what day is the last day they can call you.
http://www.ftc.gov/bcp/conline/pubs/buspubs/cal
Now, I'm not suggesting they aren't violeting the intent of the rule, but they would not be violating the rules as described on the website.
Everyone is talking about the exemptions. No-one is talking about the delay.
If you sign up before August 1st, telemarketers have to stop calling you in October. This suggests that telemarketers should/will use the list as a "call now" list during September. Similarly, if you sign up after Aug 1, they have three months to call you as many times as posible.
Yes there has to be some delay, but three months?
I'm not saying they will all do this, but I bet some will.
Given that they are collecting this list purely for the sake of giving it out to telemarketers....
As the husband of an ecologist [HOE], I must remind you that the order of living systems comes not from individuals working togeather, but from the less fit dying before they can reproduce.
Which is exactly the problem.
The thing I forgot to hit, when I slipped into "speech" mode, is that your virus scan example differs from the RCU issue in an important way. In your example MS licenses AntiV from somebody. It's a contract about the AntiV. In the RCU case IBM licenses SysV from SCO/AT&T. RCU not only isn't in the contract, it didn't exist at the time of writing. RCU isn't "derived" from SysV in the traditional sense.
So the question is more akin to "If I invent a spoiler for a Honda, does that mean Honda can stop me from mounting it on a Yugo?"
As MS has shown, what is and isn't "part" of an OS is up for interpretation. [See IE vs Mozilla] Beyond that we have the question of "Which OS?"
[ignoring historical names] IBM used SysV to make AIX. IBM wrote RCU. IBM then added RCU to AIX [not SysV]. Is RCU a "part" of SysV?
It's development didn't require SysV. [It was done as a theoretical exercise. Followed by multiple implementations.]
It's execution doesn't depend on SysV. [It runs on/in Linux, AIX and OS/400. It doesn't run on or in the SysV code they liscenced from AT&T.]
RCU was never shipped with SysV [at least not at the point they wrote it]. They didn't add something to an existing operating system. They created a new OS based on an old one. Is Internet Explorer "part" of Windows 3.1? Win95? MacOS 9?
Imagine if you modified IE to take pluggins (yes it's already been done.) MS might be able to claim ownership of the "socket" but they can't claim ownership of everything that might get plugged into it. A license for derivatives of SysV would only cover the changes to SysV which were made so that it could acomodate the RCU code. Not the RCU code.
SCO seems to be claiming that this is wrong. That they own RCU, because RCU fits in that socket. No matter the origin of RCU. This is akin to XFree claiming ownership of KDE. Or KDE claiming ownership of Karamba. This would totally destroy everything we believe about the way interoperation and extension works. Adobe would own every photoshop pluggin.
Because this is a contract issue not a statute issue, I'm being extreme. However, the concept of ownership in these cases is not trivial. The concept of derivation is under attack. If you expand it the way SCO appears to be arguing you get absurdity. Therefor, it must fail. For the sake of our livelyhoods it must be shown to fail.