This project was created by Intel to enable support for the Intel PRO/Wireless 2100 (IPW2100) mini PCI adapter. This project is intended to be a community effort as much as is possible given some working constraints (mainly, no HW documentation is available) (Emphasis mine)
So in Intel's own words, they did not release the specs, and I can't find anything on the site that says different...
I had exactly the same problem with a Xerox printer. I bought it specifically because it had 4 tanks, and a separate print head. The problem was, I have a B&W laser printer, so I only used the inkjet when I needed color. I'd go weeks without it, then try to print. Dried up tanks and clogged print heads. My cost per page was $80: $10 each for 4 new tanks and $40 for a new printhead!
Long story short, I bought a Lexmark SC1275 Color Laser on eBay for under $600. Worth every single penny. You can get the SC1270's for less than $300 on eBay. The printer can sit *forever* and print just fine when you need it to. And now that color is so cheap and easy, I print *everything* in color. It makes a difference when everything you send to a customer has at least your logo in color, and when you use really bright white (100+) paper. They notice the difference. I've received a number of compliments.
Don't like SCO taking money from Microsoft, revoke their nmap license. Did SCO take you parking stop at the store, revoke their nmap license.
I'm either being trolled (likely) or you live in a fairyland. Do you really think that business could be conducted under such conditions?
Microsoft and IBM were locked in battle over OS/2 in 1994. OS/2 is the biggest reason Microsoft signed the consent decree that allowed them to be prosecuted for antitirust violations in 1995. At that time, Microsoft was intentionally modifying Windows 3.1 specifically in such a way that it would not run under OS/2.
Yet, at the same time, IBM sold a version of OS/2 that *included* Windows 3.1. Did I mention that IBM had a source code *license* to Windows 3.1? If it really were that easy, don't you think Microsoft would have simply revoked IBM's license?
Intel *sued* AMD over creating an Intel-compatible processor. Yet, AMD continued. How could they do that? Intel had *licensed* the design of the x86 processor series to AMD (back when AMD merely did fab work for Intel). Intel acually sued and LOST. Obviously, it's not that easy to revoke a license.
As the copyright holder, you have 100% control over your code. However, the minute you execute a license over your code, you have made your choice. You can even make multiple choices on the same codebase. However, unless a specific license that you chose to execute *SPECIFICALLY INCLUDES* language that allows you to revoke or modify the license, you cannot do so. You agreed, and you're stuck.
Software licenses are not like marriage licenses: unless specifically written to include it, they do not give one party or the other the right to change their mind on a whim.
Of course, the creator can utilize different licenses on the code, and can choose not to use a license they have used previously on a *new* or *updated* product, but the old product is still licensed under the old rules. Forever, unless the license stated an expiration time. The GPL has no revocation terms, and no expiration date.
And I'm sure that they will very nicely document an API to include third party plugins to be applied to the source DVD. You know, nice plugins such as MP3 encoding/decoding, transcoding of non-encrypted VOB's to MPEG4, things like that.
Never mind that it would also be the perfect place to plug in that DeCSS code that was just ruled not to be a trade secret!:)
Notice that it says that your modifying and redistributing of the program (things prevented by copyright law but specifically allowed by GPL) signifies your acceptance of the license. Not saying you accept it, not liking it, not agreeing with it's legal framework, but merely modifing or redistributing.
IF you do either of these things, you are then forced to meet ALL terms of the GPL. I'm sure we can all agree that: 1) SCO has redistributed nmap code.
Therefore they are 2) Bound by its restrictions.
Here's the catch: How has SCO broken any of the restrictions placed upon the code by the GPL? There are only 2: 1) You must make the source available. 2) You cannot prevent others from redistributing the code on the same terms as you were given. Has SCO prevented either of these?
There is nothing in clause 4 or 5 that says that SCO must say that they think that the GPL is a legal, enforcable license. It only says that if they choose not to meet all of the terms of the license, they may not have any of the provisions of the license, and normal copyright applies.
I hope that IBM burns SCO to the ground and then crushes the ashes to powder, but the nmap people just don't have the right to "revoke" SCO's rights to GPL code. Notice that the word "revoke" is not even in the GPL. There is no provision for it. Only users that do not comply with all terms lose their license: actually, because they are not in compliance, they never had a license, and their use defaults to that of normal copyright law. However, there is no evidence that SCO is *not* in compliance.
Everybody keeps pointing this out, but what says that SCO has not accepted--and complied with--the license? There is nothing that prevents SCO from saying one thing in the IBM case (GPL is invalid), yet saying another in the Nmap case. Until a court says that one or the other is true, there is nothing that prevents them from arguing both sides. Corporations are rarely consitent, even when it would *help* them to be!:)
Nmap has claimed that SCO has violated clause 4: they lose their license because they have not obeyed the restrictions of the GPL. There are two restrictions: you must include the source, and you cannot add any additional restrictions, including on redistribution. Has SCO not included the nmap source code? Has SCO prevented anyone from redistributing the included nmap code, in source or binary? Nmap certainly is not claiming that SCO is violating either of these provisions. Therefore, SCO is arguably in compliance; therefore, they cannot have their license "revoked".
Is SCO's position here in conflict with their position vis-a-vis IBM? Certainly. Is this illegal? No. Does this mere fact prove that they are in violation of clause 4? Not until a judge says so.
Again, there is *nothing* that prevents them from accepting a license, while legally arguing that the license is invalid. Nothing.
Also, just because a contract does not expressly exclude something does not mean that the opposite is granted. So your non-exclusion does not mean an inclusion. A contract doesn't say that I can't repossess your car. It doesn't say I *can* either: so I am not granted that right.
As the copyright holder, you can put your code under 42 licenses. A *lot* of projects dual-license GPL and a proprietary-friendly license. This way, they can do *exactly* what the GCC people did: sell a proprietary copy of their license.
That has *nothing* to do with changing the license already granted to the code. It's no different from selling an item, or giving it away free with, say, a service contract. Just because you do one thing (charge for it) doesn't mean you can't do something else (give it away) under different circumstances. The two stances are incompatible, but perfectly legal.
It really boils down to this: show me where in the GPL it says that the copyright holder can alter or revoke the license? It says that, due to actions of the user the user may not be eligible to exercise the license, but there is absolutely no verbiage that says that the copyright holder can unilaterally revoke a user's rights.
So, there is only one question: is SCO in compliance with the license? The feelings of the copyright holder are meaningless, except that the copyright holder is the one that may or may not institute legal proceedings in the defense of their copyright. However, SCO does not have to agree with their interpretation: that's for a court to decide.
In other words, the nmap people saying that they're "revoking" SCO's right to use the code under clause 4 is no more or less legally binding than the GPL itself. Again, SCO is allowed to argue both sides: take advantage of the GPL with nmap, while saying that it's unconstitutional with IBM. Isn't justice grand?
By you licensing a project under the GPL and by me redistributing it, you and I have entered into a contract. Actually, you have given me terms for license, and I have agreed to follow them. You can't unilaterally change it. Licenses and contracts *cannot* be revoked if the contract doesn't say it can be. If it could be, there'd be no SCO v. IBM: SCO's case would be open-and-shut.
You can't argue both ways: either both SCO and nmap can revoke their license, or they can't. Neither contract (arguably) has a revocation clause, so the standards for one should be the same for the other.
The question is, is SCO in agreement with the license? Clause 4 says they lose their license if they do not obey the restrictions of the license. Has SCO not included the nmap source code? Has SCO prevented anyone from redistributing the nmap code, in source or binary? It is not illegal to charge for GPL code, so that is not a problem. Assuming the other items are true (and nmap is not claiming that they're not), SCO is arguably in compliance; therefore, they cannot have lost their license.
Don't get me wrong: I'd love to see SCO lose their license to *all* GPL code. I'd love to see a GPL 2.1 expressly forbidding SCO, and all GPL projects to adopt it. But it wouldn't prevent them from distributing whatever code was previously released under the GPL 2.0.
Here's the interesting thing about the law. IANAL,...
You can work under two incompatible view of the facts. While SCO can argue until its blue in the face that the GPL is invalid, it can still take advantage of the GPL until it is proven to be invalid.
There is nothing that prevents a company from fighting against a law/license/etc. and simultaneously using the exact same law to their advantage. This happens *all* the time. The DMCA comes to mind. There's nothing in the law that says a company's policy or position needs to be internally harmonious...
I do not believe this is true. If code has been licensed by the GPL, there is no mechanism for revoking that. So any code put under the GPL stays under the GPL. You can *change* the license on *new* code, but the old code remains under the old license.
Case in point: XFree86. They're changing their license to one tht Linux distributions claim is not GPL-compatible. So, Linux distributions are remaining with the older, compatible code under the older, compatible license. XFree86 can't revoke the old license; they can prevent users from getting their new code.
In other words, if 15, 2, 1, etc. does not get it to shut down cleanly, it is a buggy piece of garbage and should be permanently removed from your system.
1) MPE is a FCC standard that everybody has to follow. ARRL has as much right to complain about BPE than anyone else has to complain about AR users sitting too close too their microwave horn. In my book, it's very little, but in the FCC's book, it's a lot. The ARRL has experience with MPE, and is pointing out a problem. Yes, they don't like BPE for other reasons *TOO*, but that does not in any lessen the issue. No matter how much you wish it might.
2) Who's in the right here, the minority that already enjoys AR or the potential majority that will benefit of broadband?
If there are reasons why the new technology is going to destroy something that *every* every other provider of high-speed Internet has left alone, who's fault is that? AR users for existing? Even though you may not have a use for it, the federal government has already placed a high value on it, and the ARRL exists to protect it!
Just because you don't like the way that the FCC has prioritized does not mean you are right. Until the FCC removes the rules on MPE or removes AR as the primary allocation on the frequencies they've given them, according to 70+ years of FCC rule, BPE must be the one to make the accomodations. Period.
So amateur radio users have to comply with FCC regulations regarding MPE, but nobody else does? Why not? Why is it wrong of the ARRL to point out ways that this new technoology is or might break the rules--rules not created by the ARRL, but rather by the FCC?
The FCC has stated that the service provided by Amateur Radio are valuable enough to "the public good" to give them primary allocations of frequency. As of right now, BPE has brought exactly zero benefits to the public good. Could it? Maybe. Has it? No. And until that changes, why should BPE be allowed to break the rules that everybody else has to play by?
You know, in near 20 years of playing, I've never seen *anyone* go bankrupt to the bank. So, I did that from memory. I guess it was wrong! I didn't think that was possible!:)
I've never heard of Atlantik, but if it's a way to get competitive Monopoly players, I'm googling right now!:)
You know, I almost wrote about that. It certainly is an issue; however, I was not around amateur radio when CATV was being rolled out on a wide basis (the 70's, around here, anyway). Again, I wouldn't care if the ARRL, or XYZ group complained with legitimate worries about possible problems. If the provider of the new service (BPL, cable, whatever) can show that their fears are unfounded, fine. If not, then the new guys need to figure out how to play nice.
And that's exactly how the Amateur Radio community usually works. If you've got someone on 10 meters that's interfering with TV, it's *far* easier to build a filter for them than to say, "eat it, I've got the license." And I have a feeling that's how the *vast* majority of licensees would handle it.
But when you *do* have the license, and their horribly cheap TV is poorly (or better yet, improperly) built, with a front end that a walkie-talkie could overload, then what? I'm not buying you a new TV, and I *DO* have the FEDERAL LICENSE, and the RIGHT to the frequency I'm using!
Sombody has to have the upper hand. The government has decided that Amateur Radio has advantages such that they are willing to give us Primary use of a few narrowly defined frequencies, and Secondary use of a few more. These frequencies were not given to us. They were allocated for us, in exchange for our using them for the public good: emergency communication, etc. You may not see it as important, but the government does. Until the government, by law, changes this, that's how it works.
Do you really think the ARRL has nothing to do but complain about high speed Internet access? Have you noticed where most Amateur Radio operators are located? Not usually in the cities, where DSL and cable are cheap and plentiful... I have a feeling that most of their members would *love* cheap high-speed Internet access. As long as it doesn't destroy their *other* hobby.
I don't recall the ARRL trying to block cable Internet or DSL. Those technologies did not threaten to interfere with huge swaths of radio frequencies. I don't recall the ARLL trying to block satellite Internet access, even though there are amateur satellites. Again: they coexist peacefully. BPL, though, as yet, has not proven itself not to.
Yes, that's what tests are for. But the ARRL wants to make sure that their concerns (the concerns of their *members*) are on record. If they are addressed, then *everyone* is happy, including the ARRL members in the sticks who will benefit from both clear Amateur Radio communications as well as high-speed Internet access.
OK, this is the *definition* of off-topic. No karma bonus here!:)
When you mortgage a property, you get 1/2 of it's value from the bank, but just like real propery, you still own it. And, just like real property, you can sell it with a mortgage. The person who buys it must immediately pay the bank 10% of the mortgaged value (exactly as if they were going to unmortgage it). If they choose to unmortgage the property at that time, they may at that time *only* pay just the mortgaged value (for a grand total of the mortgaged value plus 10%, just like always). If, however, they wait, they will have to pay the normal amount.
Ever notice how weird the word mortgage is? No? Try typing it 50 times in 3 sentences...
Now, jail: If you land in jail, you may pay $50 to get out before your first, second, or third roll. At that point, you are released. If you roll doubles, you get to re-roll. If, however, you do *not* pay to get out of jail before your first, second, or third roll, you still have to roll. If you roll doubles, you get out of jail for free, but you do not get to re-roll. If after your third time you have not paid or rolled doubles, you must then pay the bank $50. If you do not have $50, you either must remove houses, mortgabe, sell a piece of property to another player, etc. If you cannot do any of these things, you become bankrupt to the bank, your remaining money (obviously less than $50) is divided among the remaining players, and your properties are then auctioned off by the bank.
I *love* Monopoly. I just don't get to play it very often, even when I tell players that in 20 years of playing monopoly, I've only had a game go beyond 3 hours twice...
Slashdot Mirror
Quote from the first page at http://ipw2100.sourceforge.net/:
This project was created by Intel to enable support for the Intel PRO/Wireless 2100 (IPW2100) mini PCI adapter. This project is intended to be a community effort as much as is possible given some working constraints (mainly, no HW documentation is available) (Emphasis mine)
So in Intel's own words, they did not release the specs, and I can't find anything on the site that says different...
Long story short, I bought a Lexmark SC1275 Color Laser on eBay for under $600. Worth every single penny. You can get the SC1270's for less than $300 on eBay. The printer can sit *forever* and print just fine when you need it to. And now that color is so cheap and easy, I print *everything* in color. It makes a difference when everything you send to a customer has at least your logo in color, and when you use really bright white (100+) paper. They notice the difference. I've received a number of compliments.
I'm either being trolled (likely) or you live in a fairyland. Do you really think that business could be conducted under such conditions?
Microsoft and IBM were locked in battle over OS/2 in 1994. OS/2 is the biggest reason Microsoft signed the consent decree that allowed them to be prosecuted for antitirust violations in 1995. At that time, Microsoft was intentionally modifying Windows 3.1 specifically in such a way that it would not run under OS/2.
Yet, at the same time, IBM sold a version of OS/2 that *included* Windows 3.1. Did I mention that IBM had a source code *license* to Windows 3.1? If it really were that easy, don't you think Microsoft would have simply revoked IBM's license?
Intel *sued* AMD over creating an Intel-compatible processor. Yet, AMD continued. How could they do that? Intel had *licensed* the design of the x86 processor series to AMD (back when AMD merely did fab work for Intel). Intel acually sued and LOST. Obviously, it's not that easy to revoke a license.
As the copyright holder, you have 100% control over your code. However, the minute you execute a license over your code, you have made your choice. You can even make multiple choices on the same codebase. However, unless a specific license that you chose to execute *SPECIFICALLY INCLUDES* language that allows you to revoke or modify the license, you cannot do so. You agreed, and you're stuck.
Software licenses are not like marriage licenses: unless specifically written to include it, they do not give one party or the other the right to change their mind on a whim.
Of course, the creator can utilize different licenses on the code, and can choose not to use a license they have used previously on a *new* or *updated* product, but the old product is still licensed under the old rules. Forever, unless the license stated an expiration time. The GPL has no revocation terms, and no expiration date.
Never mind that it would also be the perfect place to plug in that DeCSS code that was just ruled not to be a trade secret! :)
A better example may have been, say, OpenOffice.org's License strategy. They *do* dual-license for a very similar reason.
I feel so dirty...
Notice that it says that your modifying and redistributing of the program (things prevented by copyright law but specifically allowed by GPL) signifies your acceptance of the license. Not saying you accept it, not liking it, not agreeing with it's legal framework, but merely modifing or redistributing.
IF you do either of these things, you are then forced to meet ALL terms of the GPL. I'm sure we can all agree that: 1) SCO has redistributed nmap code.
Therefore they are 2) Bound by its restrictions.
Here's the catch: How has SCO broken any of the restrictions placed upon the code by the GPL? There are only 2: 1) You must make the source available. 2) You cannot prevent others from redistributing the code on the same terms as you were given. Has SCO prevented either of these?
There is nothing in clause 4 or 5 that says that SCO must say that they think that the GPL is a legal, enforcable license. It only says that if they choose not to meet all of the terms of the license, they may not have any of the provisions of the license, and normal copyright applies.
I hope that IBM burns SCO to the ground and then crushes the ashes to powder, but the nmap people just don't have the right to "revoke" SCO's rights to GPL code. Notice that the word "revoke" is not even in the GPL. There is no provision for it. Only users that do not comply with all terms lose their license: actually, because they are not in compliance, they never had a license, and their use defaults to that of normal copyright law. However, there is no evidence that SCO is *not* in compliance.
Nmap has claimed that SCO has violated clause 4: they lose their license because they have not obeyed the restrictions of the GPL. There are two restrictions: you must include the source, and you cannot add any additional restrictions, including on redistribution. Has SCO not included the nmap source code? Has SCO prevented anyone from redistributing the included nmap code, in source or binary? Nmap certainly is not claiming that SCO is violating either of these provisions. Therefore, SCO is arguably in compliance; therefore, they cannot have their license "revoked".
Is SCO's position here in conflict with their position vis-a-vis IBM? Certainly. Is this illegal? No. Does this mere fact prove that they are in violation of clause 4? Not until a judge says so.
More thoughts in this threadlet.
Also, just because a contract does not expressly exclude something does not mean that the opposite is granted. So your non-exclusion does not mean an inclusion. A contract doesn't say that I can't repossess your car. It doesn't say I *can* either: so I am not granted that right.
That has *nothing* to do with changing the license already granted to the code. It's no different from selling an item, or giving it away free with, say, a service contract. Just because you do one thing (charge for it) doesn't mean you can't do something else (give it away) under different circumstances. The two stances are incompatible, but perfectly legal.
It really boils down to this: show me where in the GPL it says that the copyright holder can alter or revoke the license? It says that, due to actions of the user the user may not be eligible to exercise the license, but there is absolutely no verbiage that says that the copyright holder can unilaterally revoke a user's rights.
So, there is only one question: is SCO in compliance with the license? The feelings of the copyright holder are meaningless, except that the copyright holder is the one that may or may not institute legal proceedings in the defense of their copyright. However, SCO does not have to agree with their interpretation: that's for a court to decide.
In other words, the nmap people saying that they're "revoking" SCO's right to use the code under clause 4 is no more or less legally binding than the GPL itself. Again, SCO is allowed to argue both sides: take advantage of the GPL with nmap, while saying that it's unconstitutional with IBM. Isn't justice grand?
By you licensing a project under the GPL and by me redistributing it, you and I have entered into a contract. Actually, you have given me terms for license, and I have agreed to follow them. You can't unilaterally change it. Licenses and contracts *cannot* be revoked if the contract doesn't say it can be. If it could be, there'd be no SCO v. IBM: SCO's case would be open-and-shut.
You can't argue both ways: either both SCO and nmap can revoke their license, or they can't. Neither contract (arguably) has a revocation clause, so the standards for one should be the same for the other.
The question is, is SCO in agreement with the license? Clause 4 says they lose their license if they do not obey the restrictions of the license. Has SCO not included the nmap source code? Has SCO prevented anyone from redistributing the nmap code, in source or binary? It is not illegal to charge for GPL code, so that is not a problem. Assuming the other items are true (and nmap is not claiming that they're not), SCO is arguably in compliance; therefore, they cannot have lost their license.
Don't get me wrong: I'd love to see SCO lose their license to *all* GPL code. I'd love to see a GPL 2.1 expressly forbidding SCO, and all GPL projects to adopt it. But it wouldn't prevent them from distributing whatever code was previously released under the GPL 2.0.
You can work under two incompatible view of the facts. While SCO can argue until its blue in the face that the GPL is invalid, it can still take advantage of the GPL until it is proven to be invalid.
There is nothing that prevents a company from fighting against a law/license/etc. and simultaneously using the exact same law to their advantage. This happens *all* the time. The DMCA comes to mind. There's nothing in the law that says a company's policy or position needs to be internally harmonious...
I do not believe this is true. If code has been licensed by the GPL, there is no mechanism for revoking that. So any code put under the GPL stays under the GPL. You can *change* the license on *new* code, but the old code remains under the old license.
Case in point: XFree86. They're changing their license to one tht Linux distributions claim is not GPL-compatible. So, Linux distributions are remaining with the older, compatible code under the older, compatible license. XFree86 can't revoke the old license; they can prevent users from getting their new code.
However, Samba certainly would hurt them. What GUI is on OpenServer?
In other words, if 15, 2, 1, etc. does not get it to shut down cleanly, it is a buggy piece of garbage and should be permanently removed from your system.
1) MPE is a FCC standard that everybody has to follow. ARRL has as much right to complain about BPE than anyone else has to complain about AR users sitting too close too their microwave horn. In my book, it's very little, but in the FCC's book, it's a lot. The ARRL has experience with MPE, and is pointing out a problem. Yes, they don't like BPE for other reasons *TOO*, but that does not in any lessen the issue. No matter how much you wish it might.
2) Who's in the right here, the minority that already enjoys AR or the potential majority that will benefit of broadband?
If there are reasons why the new technology is going to destroy something that *every* every other provider of high-speed Internet has left alone, who's fault is that? AR users for existing? Even though you may not have a use for it, the federal government has already placed a high value on it, and the ARRL exists to protect it!
Just because you don't like the way that the FCC has prioritized does not mean you are right. Until the FCC removes the rules on MPE or removes AR as the primary allocation on the frequencies they've given them, according to 70+ years of FCC rule, BPE must be the one to make the accomodations. Period.
The FCC has stated that the service provided by Amateur Radio are valuable enough to "the public good" to give them primary allocations of frequency. As of right now, BPE has brought exactly zero benefits to the public good. Could it? Maybe. Has it? No. And until that changes, why should BPE be allowed to break the rules that everybody else has to play by?
I've never heard of Atlantik, but if it's a way to get competitive Monopoly players, I'm googling right now! :)
But when you *do* have the license, and their horribly cheap TV is poorly (or better yet, improperly) built, with a front end that a walkie-talkie could overload, then what? I'm not buying you a new TV, and I *DO* have the FEDERAL LICENSE, and the RIGHT to the frequency I'm using!
Sombody has to have the upper hand. The government has decided that Amateur Radio has advantages such that they are willing to give us Primary use of a few narrowly defined frequencies, and Secondary use of a few more. These frequencies were not given to us. They were allocated for us, in exchange for our using them for the public good: emergency communication, etc. You may not see it as important, but the government does. Until the government, by law, changes this, that's how it works.
I don't recall the ARRL trying to block cable Internet or DSL. Those technologies did not threaten to interfere with huge swaths of radio frequencies. I don't recall the ARLL trying to block satellite Internet access, even though there are amateur satellites. Again: they coexist peacefully. BPL, though, as yet, has not proven itself not to.
Yes, that's what tests are for. But the ARRL wants to make sure that their concerns (the concerns of their *members*) are on record. If they are addressed, then *everyone* is happy, including the ARRL members in the sticks who will benefit from both clear Amateur Radio communications as well as high-speed Internet access.
KC8PWV
When you mortgage a property, you get 1/2 of it's value from the bank, but just like real propery, you still own it. And, just like real property, you can sell it with a mortgage. The person who buys it must immediately pay the bank 10% of the mortgaged value (exactly as if they were going to unmortgage it). If they choose to unmortgage the property at that time, they may at that time *only* pay just the mortgaged value (for a grand total of the mortgaged value plus 10%, just like always). If, however, they wait, they will have to pay the normal amount.
Ever notice how weird the word mortgage is? No? Try typing it 50 times in 3 sentences...
Now, jail: If you land in jail, you may pay $50 to get out before your first, second, or third roll. At that point, you are released. If you roll doubles, you get to re-roll. If, however, you do *not* pay to get out of jail before your first, second, or third roll, you still have to roll. If you roll doubles, you get out of jail for free, but you do not get to re-roll. If after your third time you have not paid or rolled doubles, you must then pay the bank $50. If you do not have $50, you either must remove houses, mortgabe, sell a piece of property to another player, etc. If you cannot do any of these things, you become bankrupt to the bank, your remaining money (obviously less than $50) is divided among the remaining players, and your properties are then auctioned off by the bank.
I *love* Monopoly. I just don't get to play it very often, even when I tell players that in 20 years of playing monopoly, I've only had a game go beyond 3 hours twice...