The file manager is a fancy thing that SGI cooked up for demos many moons ago. I forget the name, but I'm sure you could find it if you looked. It is an actual working file manager, though.
"really leveraging your applications look and feel" sounds to me like code for "it'll be ugly as shit and look nothing like the rest of my applications". Some people prefer things like Java or Qt because the app looks the same on all platforms. I prefer that my app look (and act) native on all platforms, which is why I use wxWindows. And it's free!
He's playing to his audience. The people who's opinion he cares about are the hardcore buisness Wall Street types ("greed is good"), who'll snap up a companies who do things like pay fines instead of complying with law cause it's cheaper. At the same time, alot of em have very John Wayne sorta mentality and they like the idea of a company coming from behing and beating the super blue-chips like IBM. He says a bunch of stuff in that artical which just barely falls short of being outright false - the way he misrepresents Perens statements about the code shown at the SCO roadshow, for example. If I hadn't been following this story, and especially if I weren't familiar with Linux, I'd be pretty heavily swayed by this kind of interview. He doesn't talk about specifics. He plays the underdog card VERY heavily. He sounds very reasonable and very straightforward. It's too bad so much of what he says is unsupportable.
Tell you what. Give me your address. I'll come to your house and pour a regular old cup of coffee on your groin. It'll hurt like a bitch and if you don't get out of your pants you'll get first and maybe second degree burns. We'll wait a couple weeks for you to heal, and then I'll pour coffee like McDonalds used to make it on you. You'll have 3rd degree burns and you may very well be permanently disfigured.
If she'd just burned herself a little, nobody would have given a shit. This was coffee that was heated to physically harmfull levels. Trying to drink it would have raised blisters on your tongue. What the hell is so hard to understand about that?
I'm going to assume that you actually aren't a programmer when you say that. For alot of us, it's a hobby as well as as a profession. Code I write on my own time is mine, not my employers. It's the same for, say, a graphic artist at a company. He works on company time designing logos or whatnot, but stuff he produces on his own time doesn't belong to his company. And while I may be happy to see it picked up by my employer and made part of a product, I might not want that too. And I should have the right to decide.
Note that you don't have any case whether you go to them or not - the argument here is that they own EVERYTHING YOU PRODUCE. Most employee argeements are worded sufficently broadly that they would also include, say, a book you wrote while employed by the company, or photos that you took. Thats why laws like these exist. You aren't (legally) owned any compensation, and thats exactly what most people get - none. They'll lose thier job if they even fight it, regardless of whether or not they win. Employers are in a really strong position in this kind of thing.
I can't even put into words how much it bothers me that you think it's reasonable to base your entire life around your job rather than drawing a seperation. I hope you insist on getting paid for 24/7 work, at least.
The price of disposing of the bodies of all the scientists you kill would be enough to feed an entire region of people in Ethepia. We'd need to kill all the funeral directors, too.
It's probably because you're pointing at a large, interesting project that gives us lots of information about the fundamental structure of matter and energy (and who could even begin to speculate how important this kind of knowledge could be in years to come), and you (at least appear to be) knocking it because it doesn't produce anything that Pfeizer can sell.
Employee protection laws exist so the companies don't exploit the workers, just like non-compete, non-disclosure, and idea-ownership clauses exist so the workers don't fuck over the companies.
Non-disclosure and non-compete contracts exist the keep the company from being screwed by subvervise employees. IP owernship contracts exist ONLY to exploit them. I can entertain, just barely, that a company should be able to keep you from releasing a product that might compete with them (of course, if I'm a carpenter and I work for a contractor, theres nothing that can keep me from moonlighting, either - IP law is fucked up). There is NO reasonable explanation for why the code should BELONG to them. Period. That, of course, is why theres law specifically stating that. The fact that your employer, practically by definition, has more finiancial and thus legal power than you doesn't change that, although it can keep you from exercising your rights.
I don't know where this rumor that you're expected to contribute positively and not effect the company negatively in any way came from. I don't hear anyone (anymore) claiming that companies are expected to contribute positively to thier employees. My responsibility to my company ends at the door. I'd probably be alot more liberal in that viewpoint if it weren't for asshat shit like this going on all the time, and if I didn't think that my job could be terminated any time the balance sheet showed that it'd be cheaper to hire an H1B. You're claiming that not obeying your contract is dishonest. I'm claiming that companies exerting pressure to enforce contract clauses that are illegal is both dishonest and reprehensible. End running around that by releasing code anonymously is perfectly moral.
"Because it sounds interesting" is the best possible reason there is to investigate something. If you only ever do research with a concrete aim in mind, you're never going to expand beyond that goal. Sounds to me like what you want to be doing is corporate R&D, not academic research.
You seem to be under the impression that contracts overrride law, which they don't. You also seem to be under the impression that a company leveraging every means at its disposal to make money is moral while a private person doing it is not. You are ALSO assuming that an employee is obligated to provide for the well being of his employer 24/7, with no commesurate obligation from the employer. On behalf of workers everywhere, I'd like to give you a big Fuck You. Employee protection laws exist for a reason.
Theres actually several tens of thousands of results for Linux. The first set of numbers are the hand-crafted (and paid for) directory results. MS doesn't show you how many actual results there are till you get past those results. They seem to only do this for certain keywords (perhaps there is some sort of threshold of paid results).
It would be a good choice if the "featured sites" weren't the paid ones. MSN not only shows you the paid listings first, they're intentionally making it difficult to see beyond them. Thats borderline unethical, imo.
Thats because they're either stupid to a degree that boggles the imagination or actively malicious and petty, using SF for free project hosting. I say fuck em. Any code that was posted on SF was released under an open source license (you have to specify the license when you create a SF project, so the fact that it's not specified in the files doesn't matter). Legal threats and posturing just make them look like little kids.
wxWindows 2.5 and later includes raw bitmap support (DIB, under windows). It was added largely because theres alot of people writing visualization software with wxWindows and there were alot of questions about wxImage and wxBitmap performance.
If you want something more cross platform than DIBs, you can try using OpenGL.
You'll be happy to know that part of the current development on wxWindows is working towards replacing (or rather, supporting) replacing the macro based container classes and home rolled string classes with wrappers and typdefs around the STL, as well as making the library exception safe. Templates are already being used to a minor degree, but even today template support is just too iffy and varies too much between compilers to really go all out - look at the hoops the Loki and Boost people need to go through to make thier libraries work.
Check out http://vcf.sourceforge.net for a library thats based on similiar concepts to wxWindows but written using modern C++ features. It's not mature yet, but maybe you can give it some help;)
Borlands VCL is actually not a wrapper around natice controls. Most of them are implemented using Win32 primitives. I have no idea why they did it this way but I assume it dates back to the early days of Dephi when the Win32 common controls were much less functional.
There's nothing wrong with that, but it does make Qt inappropriate for (and unaccessible to) the (very) large market of free, share, and low cost software on windows. Not to mention OSS projects that might want a Windows port.
It is, of course, Trolltechs right to do the licensing that way and I can even agree with thier reasons, but the simple fact is that I can't use it, and since I can't use it I'll use something else (and I've grown to prefer wxWindows anyway), and that means that when I get rich and famous or when Microsoft buys my startup or whatever, I'm going to CONTINUE using my other option. Because once I've already got a mature cross platform codebase, Qt doesn't have a whole lot to offer me.
This "hole" requires interactive logon. As a previous poster mentioned, most Linux bugs that are similiar aren't even counted.
The article you link to is a specific exploit - it's a buffer overflow in a specific, trusted process. There are alot more trusted processes on a Windows machine than on a Linux one (usually), but pointing that out would take a level of analysis that I don't think you're interested in doing. On top of that, it takes a privledged process thats running with desktop interaction, which is relatively rare. On top of THAT, designing services that aren't vulnerable to this is fairly straightforward. A similiar flaw in linux would be, say, a cgi script that called exec() on user input - it's not an insecurity in the OS that allows that to happen, but an insecurity in the process that it's implemented that way.
Theres a danger, of course, and it's something to be aware of, but it's hardly the be all and end all of fundamental security problems that you make it out to be.
Probably nothing, it's hardly a competitor for the "real" Qt. It's cygwin-hosted, not native. It's buggy. It's slow. There's been tentative work a number of times on a real, live, native win32 port of Qt/X11, but (afaik), it's never gotten very far, probably because theres other alternatives, such as (but not limited to) wxWindows.
b) wxWindows is rapidly replacing Tkinter as the standard Python GUI. There's also bindings to Perl, Ruby, Java (more than 1), JavaScript, a couple forms of Basic, Lua, and.NET. And thats just the ones I know about off the top of my head. How many of those are actively maintained, I couldn't really say, except that wxPython development parallels wxWindows.
c) I need ZERO loc to make wxWindows look like whatever platform its running on. Thats what it's for, after all.
Look, I'm not really an advocate. I like wxWindows and use it, but I'm not going to cry if you don't. I have no idea why anyone would use something like Tk when theres better alternatives, but I'm sure you have good reasons. It's all about choice, right? Just don't ask me to use any of your apps:P
Borland fucked up in a lot of ways with BCBX - if I were them, I wouldn't be selling it right now. Let me start a list...
a) It's mainly a wrapper around other (free) tools. It's a GOOD wrapper, but it's hard to justify the kind of money they want for a simple wrapper. I'd like to see some Borland-specific expertise in it for that kind of money (like a VCL->wxWindows integration library that let people use third party VCL components from withing wxWindows apps. Mmmmmmmm).
b) They pulled the wx support:P It was only available in the "technology preview", which is only available on the Enterprise version CDs. Postings in the newsgroup indicate that it'll only be available in the enterprise edition, and thats pretty damn pricy consiering that the RAD tool they show in the demo is pretty lame compared to existing wxWindows RAD tools. Again, it's a good wrapper, but I can't see dropping a grand on it - not when I can get wxDesigner for 100 bucks.
c) No code completion. This one boggles me - they've got a perfectly functional class parser already, it wouldn't have taken more than a developer-week to hack out functional code completion/insight. This is a massive stumbling block, imo.
d) No VCL migration path. I don't care about this one, but they've got a big customer base there thats really pissed off. One guy managed to get BCBX to build his old BCB6 files after some tweaking, the least they could have done is had an import function.
BCBX has all the hallmarks of an app that was a good idea but was ruined by marketing & sales. The enterprise edition is really expensive for what it provides. Basic IDE functionality is lacking. Theres nothing outstanding that you can point at as making this a unique product - almost everything it does it does as a wrapper to something I can download for free.
That said, it IS really cool. If it had code completion I'd start migrating my projects to it, even with the work required. The built in versioning is spiffy. The source control integration is excellent. The project management for dealing with all those compilers and settings is good (far better than Visual Studios *spit*. SyncEdit mode is nifty (select a block of code, start editing an identifier, all instances of that identifier will be changed as you edit. Not really anything that find/replace can't do but it just seems.... better.)
Tk's fine if thats what you're into. I prefer wxWindows because it's
a) highly supported and actively developed b) C++ (this is a minus for many people, I know) c)Not fugly. The fugly part is a big problem with alot of toolkits, especially the old ones like Tk.
In simple words: fuck thier linguistic heritage. Things die, thats how the world works. If it's truly that important, they'll preserve it themselves without protectionist laws. If you truly love something, set it free and all that.
Slashdot Mirror
The file manager is a fancy thing that SGI cooked up for demos many moons ago. I forget the name, but I'm sure you could find it if you looked. It is an actual working file manager, though.
"really leveraging your applications look and feel" sounds to me like code for "it'll be ugly as shit and look nothing like the rest of my applications". Some people prefer things like Java or Qt because the app looks the same on all platforms. I prefer that my app look (and act) native on all platforms, which is why I use wxWindows. And it's free!
He's playing to his audience. The people who's opinion he cares about are the hardcore buisness Wall Street types ("greed is good"), who'll snap up a companies who do things like pay fines instead of complying with law cause it's cheaper. At the same time, alot of em have very John Wayne sorta mentality and they like the idea of a company coming from behing and beating the super blue-chips like IBM. He says a bunch of stuff in that artical which just barely falls short of being outright false - the way he misrepresents Perens statements about the code shown at the SCO roadshow, for example. If I hadn't been following this story, and especially if I weren't familiar with Linux, I'd be pretty heavily swayed by this kind of interview. He doesn't talk about specifics. He plays the underdog card VERY heavily. He sounds very reasonable and very straightforward. It's too bad so much of what he says is unsupportable.
If she'd just burned herself a little, nobody would have given a shit. This was coffee that was heated to physically harmfull levels. Trying to drink it would have raised blisters on your tongue. What the hell is so hard to understand about that?
Note that you don't have any case whether you go to them or not - the argument here is that they own EVERYTHING YOU PRODUCE. Most employee argeements are worded sufficently broadly that they would also include, say, a book you wrote while employed by the company, or photos that you took. Thats why laws like these exist. You aren't (legally) owned any compensation, and thats exactly what most people get - none. They'll lose thier job if they even fight it, regardless of whether or not they win. Employers are in a really strong position in this kind of thing.
I can't even put into words how much it bothers me that you think it's reasonable to base your entire life around your job rather than drawing a seperation. I hope you insist on getting paid for 24/7 work, at least.
The price of disposing of the bodies of all the scientists you kill would be enough to feed an entire region of people in Ethepia. We'd need to kill all the funeral directors, too.
It's probably because you're pointing at a large, interesting project that gives us lots of information about the fundamental structure of matter and energy (and who could even begin to speculate how important this kind of knowledge could be in years to come), and you (at least appear to be) knocking it because it doesn't produce anything that Pfeizer can sell.
Non-disclosure and non-compete contracts exist the keep the company from being screwed by subvervise employees. IP owernship contracts exist ONLY to exploit them. I can entertain, just barely, that a company should be able to keep you from releasing a product that might compete with them (of course, if I'm a carpenter and I work for a contractor, theres nothing that can keep me from moonlighting, either - IP law is fucked up). There is NO reasonable explanation for why the code should BELONG to them. Period. That, of course, is why theres law specifically stating that. The fact that your employer, practically by definition, has more finiancial and thus legal power than you doesn't change that, although it can keep you from exercising your rights.
I don't know where this rumor that you're expected to contribute positively and not effect the company negatively in any way came from. I don't hear anyone (anymore) claiming that companies are expected to contribute positively to thier employees. My responsibility to my company ends at the door. I'd probably be alot more liberal in that viewpoint if it weren't for asshat shit like this going on all the time, and if I didn't think that my job could be terminated any time the balance sheet showed that it'd be cheaper to hire an H1B. You're claiming that not obeying your contract is dishonest. I'm claiming that companies exerting pressure to enforce contract clauses that are illegal is both dishonest and reprehensible. End running around that by releasing code anonymously is perfectly moral.
"Because it sounds interesting" is the best possible reason there is to investigate something. If you only ever do research with a concrete aim in mind, you're never going to expand beyond that goal. Sounds to me like what you want to be doing is corporate R&D, not academic research.
You seem to be under the impression that contracts overrride law, which they don't. You also seem to be under the impression that a company leveraging every means at its disposal to make money is moral while a private person doing it is not. You are ALSO assuming that an employee is obligated to provide for the well being of his employer 24/7, with no commesurate obligation from the employer. On behalf of workers everywhere, I'd like to give you a big Fuck You. Employee protection laws exist for a reason.
Theres actually several tens of thousands of results for Linux. The first set of numbers are the hand-crafted (and paid for) directory results. MS doesn't show you how many actual results there are till you get past those results. They seem to only do this for certain keywords (perhaps there is some sort of threshold of paid results).
It would be a good choice if the "featured sites" weren't the paid ones. MSN not only shows you the paid listings first, they're intentionally making it difficult to see beyond them. Thats borderline unethical, imo.
Thats because they're either stupid to a degree that boggles the imagination or actively malicious and petty, using SF for free project hosting. I say fuck em. Any code that was posted on SF was released under an open source license (you have to specify the license when you create a SF project, so the fact that it's not specified in the files doesn't matter). Legal threats and posturing just make them look like little kids.
If you want something more cross platform than DIBs, you can try using OpenGL.
Check out http://vcf.sourceforge.net for a library thats based on similiar concepts to wxWindows but written using modern C++ features. It's not mature yet, but maybe you can give it some help ;)
Borlands VCL is actually not a wrapper around natice controls. Most of them are implemented using Win32 primitives. I have no idea why they did it this way but I assume it dates back to the early days of Dephi when the Win32 common controls were much less functional.
It is, of course, Trolltechs right to do the licensing that way and I can even agree with thier reasons, but the simple fact is that I can't use it, and since I can't use it I'll use something else (and I've grown to prefer wxWindows anyway), and that means that when I get rich and famous or when Microsoft buys my startup or whatever, I'm going to CONTINUE using my other option. Because once I've already got a mature cross platform codebase, Qt doesn't have a whole lot to offer me.
The article you link to is a specific exploit - it's a buffer overflow in a specific, trusted process. There are alot more trusted processes on a Windows machine than on a Linux one (usually), but pointing that out would take a level of analysis that I don't think you're interested in doing. On top of that, it takes a privledged process thats running with desktop interaction, which is relatively rare. On top of THAT, designing services that aren't vulnerable to this is fairly straightforward. A similiar flaw in linux would be, say, a cgi script that called exec() on user input - it's not an insecurity in the OS that allows that to happen, but an insecurity in the process that it's implemented that way.
Theres a danger, of course, and it's something to be aware of, but it's hardly the be all and end all of fundamental security problems that you make it out to be.
Probably nothing, it's hardly a competitor for the "real" Qt. It's cygwin-hosted, not native. It's buggy. It's slow. There's been tentative work a number of times on a real, live, native win32 port of Qt/X11, but (afaik), it's never gotten very far, probably because theres other alternatives, such as (but not limited to) wxWindows.
b) wxWindows is rapidly replacing Tkinter as the standard Python GUI. There's also bindings to Perl, Ruby, Java (more than 1), JavaScript, a couple forms of Basic, Lua, and .NET. And thats just the ones I know about off the top of my head. How many of those are actively maintained, I couldn't really say, except that wxPython development parallels wxWindows.
c) I need ZERO loc to make wxWindows look like whatever platform its running on. Thats what it's for, after all.
Look, I'm not really an advocate. I like wxWindows and use it, but I'm not going to cry if you don't. I have no idea why anyone would use something like Tk when theres better alternatives, but I'm sure you have good reasons. It's all about choice, right? Just don't ask me to use any of your apps :P
a) It's mainly a wrapper around other (free) tools. It's a GOOD wrapper, but it's hard to justify the kind of money they want for a simple wrapper. I'd like to see some Borland-specific expertise in it for that kind of money (like a VCL->wxWindows integration library that let people use third party VCL components from withing wxWindows apps. Mmmmmmmm). :P It was only available in the "technology preview", which is only available on the Enterprise version CDs. Postings in the newsgroup indicate that it'll only be available in the enterprise edition, and thats pretty damn pricy consiering that the RAD tool they show in the demo is pretty lame compared to existing wxWindows RAD tools. Again, it's a good wrapper, but I can't see dropping a grand on it - not when I can get wxDesigner for 100 bucks.
b) They pulled the wx support
c) No code completion. This one boggles me - they've got a perfectly functional class parser already, it wouldn't have taken more than a developer-week to hack out functional code completion/insight. This is a massive stumbling block, imo.
d) No VCL migration path. I don't care about this one, but they've got a big customer base there thats really pissed off. One guy managed to get BCBX to build his old BCB6 files after some tweaking, the least they could have done is had an import function.
BCBX has all the hallmarks of an app that was a good idea but was ruined by marketing & sales. The enterprise edition is really expensive for what it provides. Basic IDE functionality is lacking. Theres nothing outstanding that you can point at as making this a unique product - almost everything it does it does as a wrapper to something I can download for free.
That said, it IS really cool. If it had code completion I'd start migrating my projects to it, even with the work required. The built in versioning is spiffy. The source control integration is excellent. The project management for dealing with all those compilers and settings is good (far better than Visual Studios *spit*. SyncEdit mode is nifty (select a block of code, start editing an identifier, all instances of that identifier will be changed as you edit. Not really anything that find/replace can't do but it just seems.... better.)
a) highly supported and actively developed
b) C++ (this is a minus for many people, I know)
c)Not fugly. The fugly part is a big problem with alot of toolkits, especially the old ones like Tk.
Without wanting to feed trolls, Qt isn't LGPL anywhere, and it's not free in any sense of the word on Windows. Just to be clear.
In simple words: fuck thier linguistic heritage. Things die, thats how the world works. If it's truly that important, they'll preserve it themselves without protectionist laws. If you truly love something, set it free and all that.