Absolutely not: Because Nader got so few votew, he will have to pay for reelection. Furthermore, even if he had gotten enough votes, it would be the state who would pay for it. Go back to liking it for ethical reasons!
The New Hampshire recount will really be the most interesting one. They used the Diebold machines, there was a large difference in the exit polls & the actual votes, and Kerry already won it (so right-wingers won't object to it on partisan grounds). Moderates and all concerned about evoting should watch that state closely. Partisans should also watch it, as it could be the first domino for the states where the outcome may actually change things. If anyone contests the New Hampshire recount, quietly chuckle that they don't care about voting & just care about their man. Kerry fans might protest because he did win & Bush fans might object in anticipation of those other states.
Practicaly it would be impossible to enforce without an "ET phone home" function which would move your software into the relm of spyware.
No! The GPL restricts you from modifying a program & selling it without giving out the sourcecode. The changes I'm looking for would be similarly based on a license agreement.
When you want to control use, you need a patent, not a copyright.
This is fair, but I think that "derived work isn't nearly so concretely defined as everyone assumes. Copyright law predates software. There was an interesting branch off my post which discussed "performance rights." This is one example. I think a better example would be collage. To create a collage requires the permission of the copyright holders of the works used in that collage. Even if you change the medium, such as by including long stanzas from plays in a painting. What meaning does "derived work" have for software? Can I print the source code on a T-Shirt? Can I read it out loud at a concert?
The binaries are commonly believed to be under copyright. Yet a machine produced those from my source code. Why wouldn't the plots and other machine produced outputs also be under copyright?
A mandate to submit internal changes, to me would be an amalgamation of the worst aspects of FOSS and Proprietary software.
Well you know what they say about the road to Hell! How am I supposed to sell my peers in research on F/OSS when they skeptically ask "What if some other research group takes your code, makes changes, and is then able to print more papers then us, get more funding than us, etc.?"
Sure--I believe in collaboration & F/OSS. I could be tricky & start with a GPLed base & say that I'm stuck with it. That is: yes, the idealistic values that are argued in that short piece are good. No, they don't always trump pragmatic concerns!
My estimate is that this would be in violation of the GPL as it places additional restrictions on the work, and if your work contained GPL'd code, you'd be infringing on others copyrights. If your work didn't contain GPL'd code you would OK legaly, but GPL coder's would find your license abhorent.
I agree with this & didn't even start under the pretense that my code would be GPL-compatible or OSI-approved. I know I can't make everyone happy. And I agree that GPLers would find the license abhorent. I don't care--I care more about the views of those actually in my field of research. Most don't worry about licensing. Few write GPLed code. None would care. Everyone would love to be able to start with a shared base that F/OSS provides, but people do care that the playing ground remains level!
If I hade code I told you couldn't be released under an OSI-approved license, would you say "Screw it--go proprietary. Binaries Only. Sell it. Don't share it."
I know I won't be able to release it as Free, but I would think that it would be best to release it with as much rights as possible! People did give UW a lot of flack over Pine's restrictive license. Pine is not Free software. Does that mean that Pine is, as you say, an amalgamation of the worst aspects of FOSS and Proprietary software? Are you going to email them, asking them to stop releasing the source since that is apparently bad?
Ah--thanks very much for the correction. I had followed a long discussion on the Debian list about this issue quite a while ago & people seemed to be under the impression that anyone who used your inhouse code could release it to the outside world under the GPL at any time.
I agree that the non-signature is a common concern. In fact, that point is explicitly addressed by the GPL.
I also agree that not being able to read a license before subscribing to it would be a major concern. However, I wouldn't imagine this is a deal-breaker, as long as the license could be obtained somehow. In any case, that concern wouldn't apply to the electronic software distribution commonly used by GPL-like software: the license is included whenever the software was distributed.
To this end, states should have different Election Days....perhaps ~4? One for each Tues in November.
And how exactly would this help?
To illustrate, allow me to reduce it to four states: (A)Alabama, (B)Alaska, (C)Arizona, and (D)Arkansas. Allow me to say that each state's voters have similar criteria: They want two things foremost--(1)abortions to conitnue to be legal and (2)Little restriction on gun ownership. Only two parties: the democrats support (1) and oppose (2), while the republicans oppose (1) but support (2). Assume that, knowing nothing about the makeup of Congress, the states are indifferent as to which party represents them.
Holding a single election day for all states would vote. There would be only be a 3/8 chance that there would be equal representation from both parties. It is more likely that one party will have a majority, making it more likely that either abortions are banned or more stringent gun control is imposed.
If there were two election days, it would change the voting behavior. If Alabama and Alaska vote one week & both end up putting the reps from the same party into office (there is a 1/2 chance of this), then Arizona and Arkansas would rationally elect the other party in so that balance was maintained & so abortions aren't banned & there continues to be few restrictions on gun control. If Alabama and Alaska had voted different parties in, there would only be a 1/2 chance of maintaining balanced representation between the two parties. Still, rather than being screwed by extreme party views over half (5/8) the time as in single-day elections, they'd only be screwed one quarter (2/8) of the time.
This is, of course, greatly over-simplified. But you get the point--it is better for moderates and, assuming each party is relatively equal & that they want equally to be in power and not want the other party to be in power, it doesn't disadvantage either party.
Please note it doesn't prevent the system from being in control of a single party, if that is what the country wants (and which a previous reply to my initial post argued for). If Alaska isn't a swing state, but is for the republican, it will vote the republican no matter what. This is a good thing.
That is an as-idiotic stance to take. Does "compromise" hold no meaning to you? Of course the Northern and Southern states had polarized views. This was a compromise & wasn't to the benefit of the North any more than to the South. Don't agree? Let's rephrase your post:
You do realize, of course that the reason the 3/5th person clause was put into the Constitution was to weaken the power of the Slave-free states. Representatives are apportioned by population of people with rights, and electors for presidential elections equal representatives plus senators for the state.
By making a slave count as 3/5th of a person, you have strengthened the federal power of a slave holding state.
It was a very clever move, actually, to grant slave states more federal power. If they could have gotten away with it, I'm sure the Slave states would have had a slave count as a full person.
Anyway, New Jersey and Virginia had two opposing plans. New Jersey's plan didn't get crammed down anyone's throats. Indeed, since slaves counted as more than a half a person, the compromise was closer to what the slave states wanted.
Um, you know, all branches of gov't went to one party because the VOTERS chose that.
And, if the voters would make the same decisions after they saw the final makeup of Congress, I'd say you were right. I don't think this is likely. I voted for the republican in the representative race in my region. If I knew the president, the senate, and the rest of the representatives would also be republican, I'd have changed my vote. I don't think I'm the only one. In many of the Congressional races that have occurred when the Presidential election hasn't, the composition of Congress has shifted. In the most recent case (two years ago), it shifted towards the republicans. In many others, it has shifted away from the President.
Will of the people counts no matter the result, or will of the people counts only when you like the result?
Um..I'm not saying people shouldn't vote & that vote be heard. I'm only suggesting that more information on who is already in the various branches of government may influence which way and to what strength the will of the people would lean. This isn't partisan whining--I've voted for democrats and republicans.
I don't recall massive calls for vote reform when Clinton won both terms with less than a majority popular vote,
Umm..I do. A whole bunch of people raised the fact that more people voted against him than for him & that we should have some form of runoff voting. This was particularly true when Perot managed to gain a healthy share of votes which many thought would have otherwise gone to the Republicans. I also recall talk of voting reform across the political spectrum before the election.
I do agree that many liberals are whining about reform because they didn't get what they wanted. But, just as disgusting, many conservatives are sweeping valid concerns under the rug because they got what they wanted. In some cases this is more sleezy--they had been concerned before & those concerns magically disappear if they get their way. At least those liberals who are newly concerned about voting can plead that they were ignorant of the problems before they didn't get their way!
I'll make a what-if prediction that we wouldn't be seeing this thread if Kerry had won Ohio and the presidency with less than a majority popular vote.
I'd have submitted something similar myself, regardless of how won or how much of a landslide it was. I don't know if it would have been accepted a story & it is a shame that the most enthusiastic discussions have to occur directly after an election, rather than in the more civil & less divisively partisan times leading up to the election.
Keep in mind that this is a summary of NY Times articles & not my views. Still, I will reply to you and hence reveal my thoughts on what the Times has to say.
1. While employers are required togive you two hours, this isn't enough! 2-3 hr waits aren't uncommon in many precincts (this doesn't include travel time). Reports of waits which are much longer than this aren't unheard of.
2. Lazy people won't vote early. Early voting would help decrease those excessive wait-times discussed in 1. Many states do have early voting. It is usually beneficial & the only possible argument against it that I would entertain is that it would come at a financial cost to man the polling places a few more days. I think that increasing the number of informed and proactive voters who are ale to vote is worth this minimum cost.
3. I agree with your point. However, I don't think the NYTimes author is a/. geek either. Source code accessibility need not mean the election officials themselves inspect the code. They could hire independent experts to do so. Vote tabulation is the weakest point: it represents the single point where the most votes can be added & where verification is quite difficult. The Times didn't argue that only voting machines should be opened for better scrutiny. I would also point out that a compromised machine would be a great path to get totals to the tabulators in the first place.
11. My fading credibility? (a)I didn't write the Times article, nor the summary that was posted to comp.risks. (b)How has the credibility faded? You posted points you disagreed with but, until this point, refuted no facts I presented. You can disagree with the Times author, but that doesn't disprove his credibility.
I agree that I think relatively few (if any) have stepped forward to say they were denied their right to vote as they pleased. Many more minorities were subjected to having to cast provisional ballots. Who knows if they'd be counted. A number of "dirty tricks are aimed at minority voters every year, like fliers distributed in poor neighborhoods warning that people with outstanding traffic tickets are ineligible to vote. Laws barring former felons from voting, which disproportionately disenfranchise minorities" are also a consideration.
If you think that there is no minority vote suppression, than you agree that the point is valid--just that "an end to minority vote suppression" has already been reached.
I agree that this is the most partisan point in the bunch & understand your strong reaction, but I think the Times does make excellent points.
After seeing all branches of government go to one party, I think it is careful to prevent this in the future. In a system that is so partisan, it would be best if goverment was stalled from doing anything by not giving either party all of the reigns.
To this end, states should have different Election Days. My vote for representative isn't only based on who would be best for my region, but also who would be best to keep Congress divisive.
Having 50 election days might be over-the-top. But perhaps ~4? One for each Tues in November.
Unfortunately, this would never happen--states would bicker about who would have elections first & it would confust the electorate.
1. Election day should be a holiday (rather than penalizing employees for having to take time off to vote).
2. Early voting can allow people to vote when it is convenient for them.
3. Voter-verified audit trails, source code accessibility to election officials, spot checks of code on Election Day (as is done in Nevada's slot machines!)
4. Shorter lines at the polls, standards for numbers of voting machines and poll workers.
5. Impartial election administrators, and restrictions on insiders endorsing candidates.
6. Uniform and inclusive voter registration standards.
7. Accurate and transparent voting roll purges.
8. Uniform and voter-friendly standards for counting provisional ballots.
9. Upgraded voting machines and improved ballot design.
10. Fair and uniform voter ID rules.
11. An end to minority vote suppression, disenfranchisement, harassment, dirty tricks.
12. Improved absentee ballot procedures, e.g., downloading absentee ballots from the Internet, but avoiding the ballot-by-scan/fax/e-mail with explicit loss of privacy.
Now Novell Connector, it is GPLed. It was made to work with Exchange servers (acaird said in the first post I replied to that Outlook (and hence Exchange) kept people on Windows & I did say it was for Exchange). However, Evolution need not connect to an Exchange server: it supports iCalendar and Groupware too.
Interesting reply. It was quite informative & it does look like I was expecting copyrights to solve something they can't. I am not so sure that an academic EULA that had roots in free software copyright licenses would be unobtainable or bad.
This is far from legally clear.
Well some anti-F/OSS people say the same thing about the GPL. I think that the various shrinkwrap licenses (and other EULAs & the like have had numerous expensive lawyers thrown at them too. I suppose you've reminded me that in many cases EULAs license are detrimental, but I think it is only wishful thinking to consider the GPL de-facto & shrinkwraps to be questionable.
If they are valid it is because they form part of the contract of sale of the software: an EULA rests on contract law.
Many are written for gratis software (or, at the very least, free beta-releases) too. I think many people try to attach it to the use of software (you don't see the license when you buy the product, for example). If not, then it is at least attached to the exchange/distribution rather than the "sale" for monetary gain.
In summary, under copyright law, you need a licence to release software where the copyright is owned by someone else but you don't need a licence to publish papers or give talks. Hence copyright licences can restrict releasing software but they can't directly restrict writing papers or giving talks.
Good summary. This really should have been common sense to me. To be clear: I don't really wish to restrict writing papers or giving talks. I want to further restrict the changes that can be made without needing an open release of the changed software. Nearly every F/OSS license considers the criteria to be the release of any software. Most likely because of the limits on copyright lawthat you discussed. In many cases, this is good enough. In some, such as those I described, the public would benefit if releases were required in other cases.
Most licenses are already based on the assumption that you can restrict not only distribution, but also modification of the program (See 2, 4, and 5 in the GPL). I fail to see why additional restrictions couldn't be placed in those parts of the license.
Failing that, a separate well-thought-out EULA could be beneficial. Unfortunately, it would also seem to be incompatible with a lot of F/OSS licenses: The GPL would see a EULA or NDA as additional restrictions on use. So companies are effectively prohibited from making changes to code and releasing it "in house." This is good. My point was that, in academia especially, changes ARE often made by one person in-house & aren't always distributed. This can be bad.
Ummm.. how do you figure that mechanical engineers know how to code?
Because some in academia do code. They have made intricate custon multi-scale finite element models and the like. There is F/OSS molecular dynamics, monte carlo, rendering, and other software. Until recently, CAD has been a very noticeable gap in the science/engineering programs in the F/OSS "library."
I do agree & consider Maxima spiffy. Several years ago, it wasn't GPLed & was one of the leading mathematical packages. When I suggest it to people who are used to Mathematica, they alway tell me of different features Maxima lacks. So I merely meant to provide a disclaimer.
You're correct, of course. I shouldn't have used the term "derivative works" (though some licenses do spell out what they mean by that).
You can license what people are allowed to do with your product. I see no obvious barriers to licensing them to make modifications, so long as those modifications are released whenever they use the program to do ______. Traditionally, this is "release software." But why can't it be "publish papers or present presentations based on the software?" I've definitely seen much more restrictive licenses--ones that bar you from giving negative reviews, etc.
Mathematica is under active development by some smart people. It will be difficult to catch-up. Still, Maxima is really quite good. Many of the Matlab clones (scilab, octave) are also great.
I have a similar question that has been in the back of my head for a long time. Most F/OSS licenses only consider derived works to be improvements to your program or new programs which are released with some of your IP. I wonder if there was some license that was more restrictive, that also considered the output of modified programs as publishing a derived work. If I open source great simulation software, other researchers would be allowed to make substantial changes, generate output, and publish papers without contributing the code changes to the community. Academic integrity does limit how frequently this occurs. And many are happy to have your springboard that they do collaborate with you. Also, the journals often at least require them to document their procedure so that you can eventually figure out what they did & change it yourself. But is there any legally-binding & accepted license to protect you from when the system doesn't work?
It really will save you a lot of headache. You won't be faced with the problem of some esoteric license that hasn't found OSI approval. It will also guarantee that you are A-OK with anything that is under the GPL (which will be a lot).
"Everybody uses it" isn't enough reason to choose the GPL, but it is hard to suggest anything better if you don't know your requirements.
Slashdot Mirror
Absolutely not: Because Nader got so few votew, he will have to pay for reelection. Furthermore, even if he had gotten enough votes, it would be the state who would pay for it. Go back to liking it for ethical reasons!
The New Hampshire recount will really be the most interesting one. They used the Diebold machines, there was a large difference in the exit polls & the actual votes, and Kerry already won it (so right-wingers won't object to it on partisan grounds). Moderates and all concerned about evoting should watch that state closely. Partisans should also watch it, as it could be the first domino for the states where the outcome may actually change things. If anyone contests the New Hampshire recount, quietly chuckle that they don't care about voting & just care about their man. Kerry fans might protest because he did win & Bush fans might object in anticipation of those other states.
collage. To create a collage requires the permission of the copyright holders of the works used in that collage. Even if you change the medium, such as by including long stanzas from plays in a painting. What meaning does "derived work" have for software? Can I print the source code on a T-Shirt? Can I read it out loud at a concert?
The binaries are commonly believed to be under copyright. Yet a machine produced those from my source code. Why wouldn't the plots and other machine produced outputs also be under copyright?Well you know what they say about the road to Hell! How am I supposed to sell my peers in research on F/OSS when they skeptically ask "What if some other research group takes your code, makes changes, and is then able to print more papers then us, get more funding than us, etc.?"
Sure--I believe in collaboration & F/OSS. I could be tricky & start with a GPLed base & say that I'm stuck with it. That is: yes, the idealistic values that are argued in that short piece are good. No, they don't always trump pragmatic concerns!I agree with this & didn't even start under the pretense that my code would be GPL-compatible or OSI-approved. I know I can't make everyone happy. And I agree that GPLers would find the license abhorent. I don't care--I care more about the views of those actually in my field of research. Most don't worry about licensing. Few write GPLed code. None would care. Everyone would love to be able to start with a shared base that F/OSS provides, but people do care that the playing ground remains level!
If I hade code I told you couldn't be released under an OSI-approved license, would you say "Screw it--go proprietary. Binaries Only. Sell it. Don't share it."
I know I won't be able to release it as Free, but I would think that it would be best to release it with as much rights as possible! People did give UW a lot of flack over Pine's restrictive license. Pine is not Free software. Does that mean that Pine is, as you say, an amalgamation of the worst aspects of FOSS and Proprietary software? Are you going to email them, asking them to stop releasing the source since that is apparently bad?
Ah--thanks very much for the correction. I had followed a long discussion on the Debian list about this issue quite a while ago & people seemed to be under the impression that anyone who used your inhouse code could release it to the outside world under the GPL at any time.
I agree that the non-signature is a common concern. In fact, that point is explicitly addressed by the GPL.
I also agree that not being able to read a license before subscribing to it would be a major concern. However, I wouldn't imagine this is a deal-breaker, as long as the license could be obtained somehow. In any case, that concern wouldn't apply to the electronic software distribution commonly used by GPL-like software: the license is included whenever the software was distributed.
Holding a single election day for all states would vote. There would be only be a 3/8 chance that there would be equal representation from both parties. It is more likely that one party will have a majority, making it more likely that either abortions are banned or more stringent gun control is imposed.
If there were two election days, it would change the voting behavior. If Alabama and Alaska vote one week & both end up putting the reps from the same party into office (there is a 1/2 chance of this), then Arizona and Arkansas would rationally elect the other party in so that balance was maintained & so abortions aren't banned & there continues to be few restrictions on gun control. If Alabama and Alaska had voted different parties in, there would only be a 1/2 chance of maintaining balanced representation between the two parties. Still, rather than being screwed by extreme party views over half (5/8) the time as in single-day elections, they'd only be screwed one quarter (2/8) of the time.
This is, of course, greatly over-simplified. But you get the point--it is better for moderates and, assuming each party is relatively equal & that they want equally to be in power and not want the other party to be in power, it doesn't disadvantage either party.
Please note it doesn't prevent the system from being in control of a single party, if that is what the country wants (and which a previous reply to my initial post argued for). If Alaska isn't a swing state, but is for the republican, it will vote the republican no matter what. This is a good thing.
I do agree that many liberals are whining about reform because they didn't get what they wanted. But, just as disgusting, many conservatives are sweeping valid concerns under the rug because they got what they wanted. In some cases this is more sleezy--they had been concerned before & those concerns magically disappear if they get their way. At least those liberals who are newly concerned about voting can plead that they were ignorant of the problems before they didn't get their way!I'd have submitted something similar myself, regardless of how won or how much of a landslide it was. I don't know if it would have been accepted a story & it is a shame that the most enthusiastic discussions have to occur directly after an election, rather than in the more civil & less divisively partisan times leading up to the election.
Keep in mind that this is a summary of NY Times articles & not my views. Still, I will reply to you and hence reveal my thoughts on what the Times has to say.
/. geek either. Source code accessibility need not mean the election officials themselves inspect the code. They could hire independent experts to do so. Vote tabulation is the weakest point: it represents the single point where the most votes can be added & where verification is quite difficult. The Times didn't argue that only voting machines should be opened for better scrutiny. I would also point out that a compromised machine would be a great path to get totals to the tabulators in the first place.
1. While employers are required togive you two hours, this isn't enough! 2-3 hr waits aren't uncommon in many precincts (this doesn't include travel time). Reports of waits which are much longer than this aren't unheard of.
2. Lazy people won't vote early. Early voting would help decrease those excessive wait-times discussed in 1. Many states do have early voting. It is usually beneficial & the only possible argument against it that I would entertain is that it would come at a financial cost to man the polling places a few more days. I think that increasing the number of informed and proactive voters who are ale to vote is worth this minimum cost.
3. I agree with your point. However, I don't think the NYTimes author is a
11. My fading credibility? (a)I didn't write the Times article, nor the summary that was posted to comp.risks. (b)How has the credibility faded? You posted points you disagreed with but, until this point, refuted no facts I presented. You can disagree with the Times author, but that doesn't disprove his credibility.
I agree that I think relatively few (if any) have stepped forward to say they were denied their right to vote as they pleased. Many more minorities were subjected to having to cast provisional ballots. Who knows if they'd be counted. A number of "dirty tricks are aimed at minority voters every year, like fliers distributed in poor neighborhoods warning that people with outstanding traffic tickets are ineligible to vote. Laws barring former felons from voting, which disproportionately disenfranchise minorities" are also a consideration.
If you think that there is no minority vote suppression, than you agree that the point is valid--just that "an end to minority vote suppression" has already been reached.
I agree that this is the most partisan point in the bunch & understand your strong reaction, but I think the Times does make excellent points.
Mac is pushing iCal, but it is an open standard & you just need a webdav-enabled server on your backend to use it.
Look at OpenExchange.
After seeing all branches of government go to one party, I think it is careful to prevent this in the future. In a system that is so partisan, it would be best if goverment was stalled from doing anything by not giving either party all of the reigns.
To this end, states should have different Election Days. My vote for representative isn't only based on who would be best for my region, but also who would be best to keep Congress divisive.
Having 50 election days might be over-the-top. But perhaps ~4? One for each Tues in November.
Unfortunately, this would never happen--states would bicker about who would have elections first & it would confust the electorate.
in this series of articles (free-reg-req). Summary via comp.risks:
1. Election day should be a holiday (rather than penalizing employees for having to take time off to vote).
2. Early voting can allow people to vote when it is convenient for them.
3. Voter-verified audit trails, source code accessibility to election officials, spot checks of code on Election Day (as is done in Nevada's slot machines!)
4. Shorter lines at the polls, standards for numbers of voting machines and poll workers.
5. Impartial election administrators, and restrictions on insiders endorsing candidates.
6. Uniform and inclusive voter registration standards.
7. Accurate and transparent voting roll purges.
8. Uniform and voter-friendly standards for counting provisional ballots.
9. Upgraded voting machines and improved ballot design.
10. Fair and uniform voter ID rules.
11. An end to minority vote suppression, disenfranchisement, harassment, dirty tricks.
12. Improved absentee ballot procedures, e.g., downloading absentee ballots from the Internet, but avoiding the ballot-by-scan/fax/e-mail with explicit loss of privacy.
Now Novell Connector, it is GPLed. It was made to work with Exchange servers (acaird said in the first post I replied to that Outlook (and hence Exchange) kept people on Windows & I did say it was for Exchange). However, Evolution need not connect to an Exchange server: it supports iCalendar and Groupware too.
Most licenses are already based on the assumption that you can restrict not only distribution, but also modification of the program (See 2, 4, and 5 in the GPL). I fail to see why additional restrictions couldn't be placed in those parts of the license.
Failing that, a separate well-thought-out EULA could be beneficial. Unfortunately, it would also seem to be incompatible with a lot of F/OSS licenses: The GPL would see a EULA or NDA as additional restrictions on use. So companies are effectively prohibited from making changes to code and releasing it "in house." This is good. My point was that, in academia especially, changes ARE often made by one person in-house & aren't always distributed. This can be bad.
I wouldn't disagree 5 years ago. ALSA is fantastic today, though.
There are also plenty of custom packages with Linux audio software bundled with extremely high-end, multi-channel cards.
Yes! There is a Ximian connector for MS Exchange. Collaborative scheduling is possible.
I do agree & consider Maxima spiffy. Several years ago, it wasn't GPLed & was one of the leading mathematical packages. When I suggest it to people who are used to Mathematica, they alway tell me of different features Maxima lacks. So I merely meant to provide a disclaimer.
You're correct, of course. I shouldn't have used the term "derivative works" (though some licenses do spell out what they mean by that).
You can license what people are allowed to do with your product. I see no obvious barriers to licensing them to make modifications, so long as those modifications are released whenever they use the program to do ______. Traditionally, this is "release software." But why can't it be "publish papers or present presentations based on the software?" I've definitely seen much more restrictive licenses--ones that bar you from giving negative reviews, etc.
Read the README.win32 included in the source archive. It can be made to work natively or under cygwin.
Mathematica is under active development by some smart people. It will be difficult to catch-up. Still, Maxima is really quite good. Many of the Matlab clones (scilab, octave) are also great.
Salome is one such LGPL program that is built on Open CASCADE and other F/OSS projects. Looks quite promising.
I have a similar question that has been in the back of my head for a long time. Most F/OSS licenses only consider derived works to be improvements to your program or new programs which are released with some of your IP. I wonder if there was some license that was more restrictive, that also considered the output of modified programs as publishing a derived work. If I open source great simulation software, other researchers would be allowed to make substantial changes, generate output, and publish papers without contributing the code changes to the community. Academic integrity does limit how frequently this occurs. And many are happy to have your springboard that they do collaborate with you. Also, the journals often at least require them to document their procedure so that you can eventually figure out what they did & change it yourself. But is there any legally-binding & accepted license to protect you from when the system doesn't work?
It really will save you a lot of headache. You won't be faced with the problem of some esoteric license that hasn't found OSI approval. It will also guarantee that you are A-OK with anything that is under the GPL (which will be a lot).
"Everybody uses it" isn't enough reason to choose the GPL, but it is hard to suggest anything better if you don't know your requirements.