Also, if the trademark display was impossible to "Not!" in some way, it's likely that the Trademark would break first. After all, the mandate for trademarks is to help consumers not be mislead - not to enforce artificial monopolies. If you really had to swear that your game was a legit Sega license, it'd simply be the required instruction sequence to make the machine boot. That's an argument that should trump copyright and trademark, as both require some creativity and/or non-generic usage.
I think the dodges "Not " + "copyright foo" are just extra - they wouldn't stand up anywhere else (sell a bottle of coke with a sticker that says "this is not coke"...) and are about like the "you can't read this website if you're a cop and everything on here is just innocent discussion" disclaimers you see on crack/hack websites. They're the legal equivalent of asking if someone is a cop before offering to sell him drugs.
If you stick your delicate trademark or patent in the gears of industry like this you're likely to find it's been chewed apart.
I've actually found that platform independence is a good thing, from a developer's point of view. When you test your application it gives you more real-world OS tests that tend to shake out more platform independent bugs (race conditions that die on Linux but could have gone wrong anywhere) than you'd expect. Users of other OSes make different assumptions about software - makes them a good addition to a test team.
Most of the platform specific bugs are in the file handling, display, and a few areas and usually are different APIs - it'll just fail to run if you try to do things the wrong way. Your AI is likely going to compile just fine, but die for some previously hidden logic error.
Canadians have a large number of rifles and shotguns per capita (compared to Britain for instance) but few handguns. There is a Canadian gun registry, but ownership is mostly unrestricted in this class, with easily obtained permits.
It's an attitude thing, you're just less likely to get shot in Canada than the US guns/gun-fatalities would imply. Maybe because Canada is smaller and a generation or two behind in the mega-cities race. Maybe it's a better minimum standard of living or something...?
Terrorists don't need much money. AK-47s are cheap, IEDs are cheap, suicide-bombers are cheap.
As for the security of Iraq, remember that the first time the USA went in they encouraged Iraqis to overthrow Saddam. When they left Saddam in power, many of these people were murdered. Unfortunately, just "pulling out" isn't a good solution either. It saves the USA money, but kills Iraqis who were just getting started on yet another try at building a country. Having gone in, I think the USA needs to stay there until the Iraqis in charge want them to go - that could be a long expensive time. Ideally though, Iraq would transition from US control to US-funded UN peacekeeping to Iraqi control. (Had Saddam or Iraq really been waging covert war (via Osama) against the states, I'd think the UN should have gone in, or at least should foot the cleanup bill. I don't believe Bush's justifications for the war though.)
In general, I'm in favour of UN peacekeeping even though it means imposing an unwanted military presence upon part of the population. In general, stopping genocides and such is worthwhile. Going into Afghanistan to give the women/persecuted a right to leave would have been reasonable.
Unfortunately as long as anyone is willing to kill to get their way, you have to be prepared to run, fight, surrender, or die. Rape and theft happen at all levels, personal to international, and you can as much reason with Charles Manson as with a country he leads. Preaching non-violence would be nice, but without a police force/army to protect you, it would be an invitation to those who wish to have power over others. Ghandi only succeeded because the British were too civilized to just kill thousands in cold blood - had that been the one of many other contemporary societies, he'd just have been shot.
Some people think they're better drivers than average. Of those, some are correct. It's really not hard to be safer than most - don't give in to road rage, don't use anything but stereo presets, if you're going to talk, don't dial the phone at least, don't tailgate...
Really, I feel that if anyone misuses a tool in such a way as to do something that would have killed someone, if they had been there, that their license should be revoked, because, obviously, they would have killed someone.
If you want to go practice crazy driving on empty roads and can have say that you drifted into the other lane because you were practicing extreme braking on an obviously deserted road, at least you were thinking about others. For someone who runs up onto the sidewalk at 20kph and crushes a parking meter while on the phone, we should bill them for the meter and take their license away as the result of an unsafe driving charge.
Some people really are better drivers, and some really are much worse - I know some people who I've only ever ridden with once...
I see no reason to give the idiots infinite chances until they kill someone, when it should be immediately obvious that they don't understand the task at hand and will eventually fail.
Wah. Novell sold out to Microsoft for $400M and someone said something that could perhaps be taken as a little presumptuous.
You're ignoring that many people honestly believe this to be like SCO, an attempt for Microsoft to kill the whole industry. At that, even if you don't see that outcome as likely, many people do and you should be able to see how in the light of this a simple position statement and invitation to the developers involved is reasonable. If you don't like it, comment to Novell about their deal
You people who blow up about having to read someone else's unsolicited opinion are babies. If you do something, expect others to comment, especially if you sell them out.
And that's really only relevant if you're doing it for a grade. If I could replace a week-long research project for one of my clients by simply pointing to a URL (quoting) and writing a paragraph or two describing the relevance to them they'd love me.
And as for school, I've long thought that we should make direct copy&paste legal in essays, if attributed (ie, as a quote), and simply give more marks for original content. This way many people who don't have the skills for original composition would still learn how to write an essay or business case because they'd have gotten some marks for what works in the real world, rather than a failing grade for ivory tower reasons.
I'm against Windows because it's almost impossible to install and use it safely unless you're a tech. I'd rather do a little manual config of Ubuntu over VNC than try to lock down my mom's WinXP box so that she can browse safely.
The plugins don't transfer between versions, but I made a list of my extensions on my website and now I only have to "Trust" my own page and click a bunch of download links.
It's not totally automatic, but versus finding and trusting a bunch of stuff, it's pretty close. The settings and history I copy around and haven't had a problem with, yet...
You assume that whatever word processor they rebranded is better than OO, such that people using it will never have problems, valid or otherwise. I imagine that the same tech-support would suffice for both - they aren't going to try to recompile anything, just check a couple obvious things and issue a refund.
I don't understand why it's supposedly going to fail so much more and need more support than the absolute nothing you get with all commercial software.
You say we should complain to the FCC, but that's not reasonable. Many of "us" are outside the USA, not English-speaking, or are otherwise unlikely to sway the opinion of a notoriously conservative government agency.
Instead, you force the equipment vendors to deliver the message.
The bottom line is that skilled hackers can crack anything and anyone can download their cracked drivers. Hiding this doesn't prevent criminal use, it prevents all the legal uses.
If the browsers are stable, etc, then I don't see what the problem is about them not being identical. If anything, being compiled in multiple code-bases should only make the project more stable for everyone. I use many versions of Firefox between work and home and they all use the same bookmarks file, the same bookmarklet scripts, the same plugins (which auto-detect which version to install for)...
It would be a problem switching between Firefox and IE though... As things like bookmarklets and plugins overwhelm the difference between IE and Firefox, etc, the cost of switching and thus the annoyance of multiple platforms will be gone.
You have to sue for damages, at that, you can get anything you convince a judge is reasonable essentially.
This slapdash approach to patent enforcement means that you really can claim anything as infringing. This is how patents are used as weapons, you tell someone that their email client, or Linux distro, will be infringing and that you'll sue them if they release it. They don't know what a judge will say, only that their lawyer estimates $1.5 million to find out, so they go along with your patent, even if it's totally unrelated and they could prove it.
Patents really are broken. They make so many assumptions about duration, enforcement, general applicability, and also are awarded to the first person to file, not the inventor, so they end up being essentially useless, unless you either plan to attack someone with them, or defend against someone else's attacks. (You'd threaten to sue the agressors over one of your patents, unless they back down.) For society, outside of a few possible areas, patents are a net loss. They're supposed to even the field, let the little guy invent someone and keep it exclusive, but with the way the courts are, this merely ends up handing *every* patent case to the richest legal team.
The essential problem at the root of it is that North American courts are broken, they're a competition between lawyers rather than a judge-directed attempt to discover the facts.
Many people loved school, myself I despised it. I can't abide by arbitrary rules and that's pretty much all schools are about. Censoring students outside of school, withholding graduation for "rules" violations, etc.
So I got a job and it was the same. Now I work for myself. My clients can still be jerks, but I get five times the pay and can ditch them if they're too bad. That takes the sting out of it.:)
Look forward to the open-textbook project and similar things meaning the end of the "doorway to business" type of schools - the fat and useless ones that abuse your for four years and $50k, just to allow you to not flip burgers.
The main reason I like the debian way is because I can get someone else to do the base install, from across the country if needed, and finish the customizations after install. I hear with yum and such, you can actually do the equivalent of 'apt-get install kde' and expect it to work. When I quit using RH and Mandrake, this could only reliably be done at install.
Otherwise, you're right, potayto, potahto.
Is it at the 'install kde' stage, or does this sort of whole-system affecting stuff still require babysitting?
The PTO's day-to-day operations are self-funded, but their government operations, legal costs, etc are not, as far as I can tell.
As for their public meetings, if you really think they're going to listen to anything other than "more patents, more money" you're naive. They are funded largely by their patent-anything operations and I highly doubt they'd give them up, especially when the organization has officially stated that they see the solution to patent disputes in court, not in a more thorough review process. Also, they are completely shielded from the legal consequences of their actions, which gives them a really poor incentive to improve.
Yes, I think software patents are useless, but from my business experience I don't think there's much value to the system in general, except perhaps for impressing shareholders with your patents. They don't help the little guy, as intended, because there are so many ways to keep a valid decision from being reached in court until your poorer opponent is bankrupt. I've worked for companies on both sides of this, and seen in in many more public cases. Patents are a government-granted monopoly with which you can beat your competition to death, but aren't very useful in compelling royalty payments. (That whole enforced in court thing, where the largest legal budget wins.)
For the overall value of the system, I'd say it's negative. In some industries, particularly those burdened by government regulations (drug industry, etc) where many of the costs are external, I think they can help. Almost everywhere else they're either a stick with which to abuse the competition, or a way to bar entry to a whole class of technologies via non-specific patents.
"Get rid of it" is a valid answer imho. It was broken at implementation, it got worse with excessive corporate pressure, and it's killing innovation now.
Of course, you immediately resort to ad-hominem attacks, it's unlikely you'll read this with an open mind...
The money already does come from us, the end users and taxpayers. We pay for the USPTO, but have absolutely zero say over anything related to it, and we pay the most terrible cost - much higher R&D costs rolled into the product prices - because of all the patent nonsense. Perhaps if we payed for a patent system that worked it'd be much cheaper in the long run. As it is, the patent system and *everyone* who works for it is a thief. They knowingly produce something of less than zero value and are totally unaccountable for it - you can't even sue them for their total fuckups without permission of the government - something you'll NEVER get. I say we go scientology on their ass - sue everyone who works at the USPTO for any little civil thing you can, waste their time and money, and stop immediately when they quit. Stick the jerks who do make money from this with ever-mounting legal bills and maybe they'll suddenly decide that the system really is a joke.
That supposedly is patent-worthy? How many ways are there? They're describing a width-first enumeration. WOW. Someone tell Knuth, he should write another volume once he figures out how all this works!!!1!
Why didn't RIM just play Scientology and sue everyone who worked for that company for everything they could, harass them out of business while stalling on the patent case?
Maybe we should have an anti-patent group who sues CEOs of abusive companies (Rambus, SCO) for a million little things. The death of a thousand small-claims cases.
Validation should happen for almost free. A good software test involves some sort of input faking and screen scraping, so as not to interfere with the running software on the machine. As such, it should be pretty simple to rerun the test on any other platform. If it's not key-for-key identical, you've failed anyways. Developing the test costs, rerunning it should happen in an automated fashion for every check-in to cvs and simply mail results to the developer.
Seriously, browser based stuff should run on almost anything, and should be tested on it. The whole point is to remove the client-system from the usage requirements.
Seems like more proof that the company is a Rambus. All patents, no solid, useful, inovative products.
Actually, I think that package selection after installation is better. It's easier and faster to get a base system running, such that you could ssh in and finish the install remotely. It's also more consistent. If you want to install Apache during install, it's not the same process as later. In the after-install model, it's always the same command.
I went through years of configuring a Redhat system and saving configs on disk. Then I went to Debian and now my system config is a few apt-get lines I cut-and-paste into the terminal after install.
You're in a GPL thread, and say you like the GPL v2, so supposedly you agree with the general principle, which is that the code is available for everyone who agrees to share equally openly.
I, as a programmer who has released GPLed code, like this. I'm not going to make any money if anyone uses anything I've written, but at least I get to know that I'm not helping some jerk who wouldn't help me.
Except, that this obligation isn't really binding. Through deceptive practices the company can appear to release the full source but avoid actually providing anything of value. This would annoy me if it happened, because imho a company that acts like a jerk shouldn't get access to my code. That's the whole point of the GPL... Reward cooperation to encourage it.
Even if I really thought enough people would hear about the GPL and care about abuse of free software to put the printer company out of business, I don't care. If I had wanted any jerk to be able to use my code, I'd have BSDLed it.
Your arguments seem just as (in)valid against the GPL vs BSDL altogether - "don't enforce it with a license, let the market handle it". I didn't buy them in that case, and I don't now.
There are many silver bullets. Go into the average C shop and give them Python or Ruby. Watch how much quicker all the "mundane" stuff gets done. Some critical piece of OS code might still need C, and they'd probably use it for glue, but regexes, complex data representations, networking, and many other things would fly.
What there aren't, is one-size-fits-all silver bullets. What's revolutionary to someone else may be mundane, or irrelevant to you. But as far as single changes that can more than double productivity (if used properly), there are many.
That said, most projects that I've seen have failed because of implementation issues... Teams who use the numeric primary key to look up the item name, then use string comparisons to check it against others, rather than using the guaranteed unique primary key... They could have used an optimized boyer-moore string search to speed this up, but it wouldn't fix the design. For projects/teams like this, there is no magic technology because whatever it is will be misused.
Slashdot Mirror
Also, if the trademark display was impossible to "Not!" in some way, it's likely that the Trademark would break first. After all, the mandate for trademarks is to help consumers not be mislead - not to enforce artificial monopolies. If you really had to swear that your game was a legit Sega license, it'd simply be the required instruction sequence to make the machine boot. That's an argument that should trump copyright and trademark, as both require some creativity and/or non-generic usage.
...) and are about like the "you can't read this website if you're a cop and everything on here is just innocent discussion" disclaimers you see on crack/hack websites. They're the legal equivalent of asking if someone is a cop before offering to sell him drugs.
I think the dodges "Not " + "copyright foo" are just extra - they wouldn't stand up anywhere else (sell a bottle of coke with a sticker that says "this is not coke"
If you stick your delicate trademark or patent in the gears of industry like this you're likely to find it's been chewed apart.
I've actually found that platform independence is a good thing, from a developer's point of view. When you test your application it gives you more real-world OS tests that tend to shake out more platform independent bugs (race conditions that die on Linux but could have gone wrong anywhere) than you'd expect. Users of other OSes make different assumptions about software - makes them a good addition to a test team.
Most of the platform specific bugs are in the file handling, display, and a few areas and usually are different APIs - it'll just fail to run if you try to do things the wrong way. Your AI is likely going to compile just fine, but die for some previously hidden logic error.
Canadians have a large number of rifles and shotguns per capita (compared to Britain for instance) but few handguns. There is a Canadian gun registry, but ownership is mostly unrestricted in this class, with easily obtained permits.
It's an attitude thing, you're just less likely to get shot in Canada than the US guns/gun-fatalities would imply. Maybe because Canada is smaller and a generation or two behind in the mega-cities race. Maybe it's a better minimum standard of living or something...?
Terrorists don't need much money. AK-47s are cheap, IEDs are cheap, suicide-bombers are cheap.
As for the security of Iraq, remember that the first time the USA went in they encouraged Iraqis to overthrow Saddam. When they left Saddam in power, many of these people were murdered. Unfortunately, just "pulling out" isn't a good solution either. It saves the USA money, but kills Iraqis who were just getting started on yet another try at building a country. Having gone in, I think the USA needs to stay there until the Iraqis in charge want them to go - that could be a long expensive time. Ideally though, Iraq would transition from US control to US-funded UN peacekeeping to Iraqi control. (Had Saddam or Iraq really been waging covert war (via Osama) against the states, I'd think the UN should have gone in, or at least should foot the cleanup bill. I don't believe Bush's justifications for the war though.)
In general, I'm in favour of UN peacekeeping even though it means imposing an unwanted military presence upon part of the population. In general, stopping genocides and such is worthwhile. Going into Afghanistan to give the women/persecuted a right to leave would have been reasonable.
Unfortunately as long as anyone is willing to kill to get their way, you have to be prepared to run, fight, surrender, or die. Rape and theft happen at all levels, personal to international, and you can as much reason with Charles Manson as with a country he leads. Preaching non-violence would be nice, but without a police force/army to protect you, it would be an invitation to those who wish to have power over others. Ghandi only succeeded because the British were too civilized to just kill thousands in cold blood - had that been the one of many other contemporary societies, he'd just have been shot.
Some people think they're better drivers than average. Of those, some are correct. It's really not hard to be safer than most - don't give in to road rage, don't use anything but stereo presets, if you're going to talk, don't dial the phone at least, don't tailgate...
Really, I feel that if anyone misuses a tool in such a way as to do something that would have killed someone, if they had been there, that their license should be revoked, because, obviously, they would have killed someone.
If you want to go practice crazy driving on empty roads and can have say that you drifted into the other lane because you were practicing extreme braking on an obviously deserted road, at least you were thinking about others. For someone who runs up onto the sidewalk at 20kph and crushes a parking meter while on the phone, we should bill them for the meter and take their license away as the result of an unsafe driving charge.
Some people really are better drivers, and some really are much worse - I know some people who I've only ever ridden with once...
I see no reason to give the idiots infinite chances until they kill someone, when it should be immediately obvious that they don't understand the task at hand and will eventually fail.
Wah. Novell sold out to Microsoft for $400M and someone said something that could perhaps be taken as a little presumptuous.
You're ignoring that many people honestly believe this to be like SCO, an attempt for Microsoft to kill the whole industry. At that, even if you don't see that outcome as likely, many people do and you should be able to see how in the light of this a simple position statement and invitation to the developers involved is reasonable. If you don't like it, comment to Novell about their deal
You people who blow up about having to read someone else's unsolicited opinion are babies. If you do something, expect others to comment, especially if you sell them out.
And that's really only relevant if you're doing it for a grade. If I could replace a week-long research project for one of my clients by simply pointing to a URL (quoting) and writing a paragraph or two describing the relevance to them they'd love me.
And as for school, I've long thought that we should make direct copy&paste legal in essays, if attributed (ie, as a quote), and simply give more marks for original content. This way many people who don't have the skills for original composition would still learn how to write an essay or business case because they'd have gotten some marks for what works in the real world, rather than a failing grade for ivory tower reasons.
I'm against Windows because it's almost impossible to install and use it safely unless you're a tech. I'd rather do a little manual config of Ubuntu over VNC than try to lock down my mom's WinXP box so that she can browse safely.
The plugins don't transfer between versions, but I made a list of my extensions on my website and now I only have to "Trust" my own page and click a bunch of download links.
It's not totally automatic, but versus finding and trusting a bunch of stuff, it's pretty close. The settings and history I copy around and haven't had a problem with, yet...
You assume that whatever word processor they rebranded is better than OO, such that people using it will never have problems, valid or otherwise. I imagine that the same tech-support would suffice for both - they aren't going to try to recompile anything, just check a couple obvious things and issue a refund.
I don't understand why it's supposedly going to fail so much more and need more support than the absolute nothing you get with all commercial software.
You say we should complain to the FCC, but that's not reasonable. Many of "us" are outside the USA, not English-speaking, or are otherwise unlikely to sway the opinion of a notoriously conservative government agency.
Instead, you force the equipment vendors to deliver the message.
The bottom line is that skilled hackers can crack anything and anyone can download their cracked drivers. Hiding this doesn't prevent criminal use, it prevents all the legal uses.
If the browsers are stable, etc, then I don't see what the problem is about them not being identical. If anything, being compiled in multiple code-bases should only make the project more stable for everyone. I use many versions of Firefox between work and home and they all use the same bookmarks file, the same bookmarklet scripts, the same plugins (which auto-detect which version to install for)...
It would be a problem switching between Firefox and IE though... As things like bookmarklets and plugins overwhelm the difference between IE and Firefox, etc, the cost of switching and thus the annoyance of multiple platforms will be gone.
You have to sue for damages, at that, you can get anything you convince a judge is reasonable essentially.
This slapdash approach to patent enforcement means that you really can claim anything as infringing. This is how patents are used as weapons, you tell someone that their email client, or Linux distro, will be infringing and that you'll sue them if they release it. They don't know what a judge will say, only that their lawyer estimates $1.5 million to find out, so they go along with your patent, even if it's totally unrelated and they could prove it.
Patents really are broken. They make so many assumptions about duration, enforcement, general applicability, and also are awarded to the first person to file, not the inventor, so they end up being essentially useless, unless you either plan to attack someone with them, or defend against someone else's attacks. (You'd threaten to sue the agressors over one of your patents, unless they back down.) For society, outside of a few possible areas, patents are a net loss. They're supposed to even the field, let the little guy invent someone and keep it exclusive, but with the way the courts are, this merely ends up handing *every* patent case to the richest legal team.
The essential problem at the root of it is that North American courts are broken, they're a competition between lawyers rather than a judge-directed attempt to discover the facts.
It doesn't seem to be. Abusive lawsuits rarely fail, as their goal isn't to win, just to suck money from the opponent.
Many people loved school, myself I despised it. I can't abide by arbitrary rules and that's pretty much all schools are about. Censoring students outside of school, withholding graduation for "rules" violations, etc.
:)
So I got a job and it was the same. Now I work for myself. My clients can still be jerks, but I get five times the pay and can ditch them if they're too bad. That takes the sting out of it.
Look forward to the open-textbook project and similar things meaning the end of the "doorway to business" type of schools - the fat and useless ones that abuse your for four years and $50k, just to allow you to not flip burgers.
The main reason I like the debian way is because I can get someone else to do the base install, from across the country if needed, and finish the customizations after install. I hear with yum and such, you can actually do the equivalent of 'apt-get install kde' and expect it to work. When I quit using RH and Mandrake, this could only reliably be done at install.
Otherwise, you're right, potayto, potahto.
Is it at the 'install kde' stage, or does this sort of whole-system affecting stuff still require babysitting?
The PTO's day-to-day operations are self-funded, but their government operations, legal costs, etc are not, as far as I can tell.
As for their public meetings, if you really think they're going to listen to anything other than "more patents, more money" you're naive. They are funded largely by their patent-anything operations and I highly doubt they'd give them up, especially when the organization has officially stated that they see the solution to patent disputes in court, not in a more thorough review process. Also, they are completely shielded from the legal consequences of their actions, which gives them a really poor incentive to improve.
Yes, I think software patents are useless, but from my business experience I don't think there's much value to the system in general, except perhaps for impressing shareholders with your patents. They don't help the little guy, as intended, because there are so many ways to keep a valid decision from being reached in court until your poorer opponent is bankrupt. I've worked for companies on both sides of this, and seen in in many more public cases. Patents are a government-granted monopoly with which you can beat your competition to death, but aren't very useful in compelling royalty payments. (That whole enforced in court thing, where the largest legal budget wins.)
For the overall value of the system, I'd say it's negative. In some industries, particularly those burdened by government regulations (drug industry, etc) where many of the costs are external, I think they can help. Almost everywhere else they're either a stick with which to abuse the competition, or a way to bar entry to a whole class of technologies via non-specific patents.
"Get rid of it" is a valid answer imho. It was broken at implementation, it got worse with excessive corporate pressure, and it's killing innovation now.
Of course, you immediately resort to ad-hominem attacks, it's unlikely you'll read this with an open mind...
The money already does come from us, the end users and taxpayers. We pay for the USPTO, but have absolutely zero say over anything related to it, and we pay the most terrible cost - much higher R&D costs rolled into the product prices - because of all the patent nonsense. Perhaps if we payed for a patent system that worked it'd be much cheaper in the long run. As it is, the patent system and *everyone* who works for it is a thief. They knowingly produce something of less than zero value and are totally unaccountable for it - you can't even sue them for their total fuckups without permission of the government - something you'll NEVER get. I say we go scientology on their ass - sue everyone who works at the USPTO for any little civil thing you can, waste their time and money, and stop immediately when they quit. Stick the jerks who do make money from this with ever-mounting legal bills and maybe they'll suddenly decide that the system really is a joke.
That supposedly is patent-worthy? How many ways are there? They're describing a width-first enumeration. WOW. Someone tell Knuth, he should write another volume once he figures out how all this works!!!1!
Why didn't RIM just play Scientology and sue everyone who worked for that company for everything they could, harass them out of business while stalling on the patent case?
Maybe we should have an anti-patent group who sues CEOs of abusive companies (Rambus, SCO) for a million little things. The death of a thousand small-claims cases.
I find it interesting that you're scared to name the school you pay, for fear they'll stop proving the service you pay for.
You're the customer. You also own the copyright to everything you've written, even if they claim otherwise.
Validation should happen for almost free. A good software test involves some sort of input faking and screen scraping, so as not to interfere with the running software on the machine. As such, it should be pretty simple to rerun the test on any other platform. If it's not key-for-key identical, you've failed anyways. Developing the test costs, rerunning it should happen in an automated fashion for every check-in to cvs and simply mail results to the developer.
Seriously, browser based stuff should run on almost anything, and should be tested on it. The whole point is to remove the client-system from the usage requirements.
Seems like more proof that the company is a Rambus. All patents, no solid, useful, inovative products.
- WNight
Actually, I think that package selection after installation is better. It's easier and faster to get a base system running, such that you could ssh in and finish the install remotely. It's also more consistent. If you want to install Apache during install, it's not the same process as later. In the after-install model, it's always the same command.
I went through years of configuring a Redhat system and saving configs on disk. Then I went to Debian and now my system config is a few apt-get lines I cut-and-paste into the terminal after install.
Let's take this step by step.
You're in a GPL thread, and say you like the GPL v2, so supposedly you agree with the general principle, which is that the code is available for everyone who agrees to share equally openly.
I, as a programmer who has released GPLed code, like this. I'm not going to make any money if anyone uses anything I've written, but at least I get to know that I'm not helping some jerk who wouldn't help me.
Except, that this obligation isn't really binding. Through deceptive practices the company can appear to release the full source but avoid actually providing anything of value. This would annoy me if it happened, because imho a company that acts like a jerk shouldn't get access to my code. That's the whole point of the GPL... Reward cooperation to encourage it.
Even if I really thought enough people would hear about the GPL and care about abuse of free software to put the printer company out of business, I don't care. If I had wanted any jerk to be able to use my code, I'd have BSDLed it.
Your arguments seem just as (in)valid against the GPL vs BSDL altogether - "don't enforce it with a license, let the market handle it". I didn't buy them in that case, and I don't now.
There are many silver bullets. Go into the average C shop and give them Python or Ruby. Watch how much quicker all the "mundane" stuff gets done. Some critical piece of OS code might still need C, and they'd probably use it for glue, but regexes, complex data representations, networking, and many other things would fly.
What there aren't, is one-size-fits-all silver bullets. What's revolutionary to someone else may be mundane, or irrelevant to you. But as far as single changes that can more than double productivity (if used properly), there are many.
That said, most projects that I've seen have failed because of implementation issues... Teams who use the numeric primary key to look up the item name, then use string comparisons to check it against others, rather than using the guaranteed unique primary key... They could have used an optimized boyer-moore string search to speed this up, but it wouldn't fix the design. For projects/teams like this, there is no magic technology because whatever it is will be misused.