People seem to think that I'm planning on taking their hard work and running away with it. I'm not. The free version and the commercial version are the same toolkit. If someone contributed to my GPL project, I wouldn't be able to close it later. I don't want to steal anyone's hard-earned work. I just want the option to switch to the commercial library later.
It's a game editor. It doesn't have much value outside the game. It does have value with the game, and if I were to release it it'd be as a free addon - however, since it uses a good chunk of the game code internally, and I don't want to open-source the game, I don't want to be forced to open-source the editor also. Therefore a GPL library isn't practical.
If they were GPL-only, I'd say, sure, our philosophies differ, and I'd move on. But they aren't. They're GPL, and commercial - but for some reason they demand I pay now if there is ever a possibility that I would want to use their commercial package. That's weird.
"Just do it, there's no way they could know" is not an acceptable option, IMHO. If they wanted to avoid the many-developers one-license thing, they could simply require one license for each person who had worked on the codebase before the conversion. I'd be fine with that. But they didn't.
As it is, I'm using wxWidgets instead. wxWidgets is basically the LGPL with an added exception, which makes it both a free software library and entirely practical for proprietary software (even more so than LGPL, in fact.)
Free software and commercial software aren't mutually exclusive, and QT is neatly killing an entire segment of the market for themselves (namely, people who later decide they want to close their software). Their choice, but I have to say it's a weird decision.
I think you're missing my point somewhat - I can't, as a small developer who doesn't even know if his software is going to be released commercially, start coding now and then purchase a license later. I'm a small game developer and my editor may be of no interest to anyone but me. But if it does turn out to be useful to release it, and I don't want to release it open-source, I can't simply buy a commercial license and be done with it.
Why should Trolltech mind if I bought a license later rather than sooner? They're still getting the license. One way just forces me to decide much earlier, when I may simply not have the information that I need to determine which is the right course of action. (Which, in this case, turns out to be "don't use QT".)
QT has two different licenses - their open-source license, and their commercial license. This is not a problem - I'm fine with this.
The problem I have is that they require that any software written for their commercial-license library be only written for their commercial-license library. This means that if, like me, you're someone trying to start a game studio looking for a basic windowing library for an editor, you have three basic choices:
* Write your editor with their free library, then never be able to distribute it in any way without GPL'ing it * Shell out $$$ for the commercial library, whether or not you'll ever need to distribute it * Use a different library
Obviously I've chosen #3, but I can't help but think that perhaps Trolltech lost a sale there - I probably would have used QT if it had been a viable option, and if I'd ever decided to distribute the editor I likely would have gladly paid the licensing fee. It's a bizarre licensing decision.
And it works, you fall for it too. How else do you know it was a SHELL gas station? If you were imune to it and not a sheep you would just tank at any gas station. (but without any advertising whatsoever, how would you know it is a gas station?) You obviously saw Shells adversting, yes even the sign that says Shell is part of advertising.
I did the same thing that the GP did, and the only reason I know it was a Shell station is because I explicitly checked once the ads started so I'd know which gas stations to avoid in the future. I wouldn't have known it was Shell if they hadn't made me care.
Except I was, actually, willing to pay. I donate pretty frequently to open-source projects. And it may have been hard the first time they did it, but considering that they were already providing a crippled win32 version I kind of doubt the "registered" version was any harder than clicking the "make it not suck" checkbox.
I suppose it's possible that they would only have ported it to Windows if people were willing to pay - but considering that there was a third party perfectly willing to make a Windows version on its own, I don't actually believe that either.
X-Chat is a very good IRC program. I use it on Windows. I was looking for an open-source chat program, I found X-Chat, and decided I liked it.
I didn't like their philosophy. They claimed that making a Windows build was "hard" somehow, but anyone who knows anything about programming knows that once you have the dev environment set up properly, making a new build is a matter of minutes. They'd obviously set up the dev environment properly, so why were they charging money for the Windows version?
However, there's somebody who makes Windows builds of X-Chat. For free. Unaffiliated with the actual group, of course, but fully functional. So I downloaded from him . . . and donated $10 to him for providing a useful service without demanding money.
Still haven't given a cent to the "main team" and I don't plan to until they get rid of that braindamaged policy.
If you split your userbase among bizarre lines, and demand money from one group, you may very well piss off the people who you're demanding money from. Be careful about things like this.
A month and a half ago my fiancee dumped me. I said, essentially, "Fuck! That sucks. Okay I'm going to slack for a while." It's been a month and a half since, and I've beaten a fuckton of video games. I'm feeling a lot better now. I think once I'm done with Super Paper Mario and the Sam and Max series I'll be ready to go back to work.
I've got enough money that I can do this. I've got enough skills that I have no worries about my ability to find a new job. I live in a 900-square-foot apartment four blocks from one of the Bay Area downtown nodes, and I have a 60" TV, no cable, all the game consoles I want, and two cars, the newest of which is an '88 Sentra without a working stereo.
I'm living my life how I want to. I have the things I care for, and I don't spend money on status symbols. I keep enough reserve that I can take a break when I need it.
I've got the "something", and I can tell you, quite honestly, "something" can be totally awesome.
I've never had my DS, PSP, XBox 360, PS2, Wii, or Gamecube crash. Ever. And I've used each one of those systems more than I've ever used a graphing calculator . . . and curiously enough, I have had my TI-81 crash. Lost all its saved data too.
Modern technology doesn't imply frequent crashing. Modern technology and complex code doesn't even imply frequent crashing. I have no doubt that you could build a cheap graphing calculator on a 400mhz XScale chip, with maybe 128mb of storage, with a full GUI and a hierarchical filesystem, just as stable as they used to be.
I would agree with your sentiments, but many Internet Service Providers are putting in capacity limits on how much you can download per month, and that will kibosh any idea of a VC-1 or AVC encoded 1080p 48 fields/second movie.
Yes, this will be an issue. I remember some of the early ISPs with 50mb limits, and those limits really haven't been raised at all since.
Honestly, I don't have that fast of a connection, and it takes me about half an hour to download an entire CD. 600 megabytes. That 200-meg file? 10 minutes. You are living in the past, and not realizing that connections will keep getting faster.
It's highly doubtful that he'd ever accept v3-only submissions since that would mean that the kernel could no longer be distributed under gplv2 *or* gplv3, but instead different parts of it would have to be distributed under different licenses. I'm not even sure that's legally possible.
I imagine he'd accept dual-licensed submissions but it's quite possible that he wouldn't bother to document the gplv3 part.
Good point, I'd forgotten about that - some people at my high school computer lab used to run SimAnt over the network (which was, obviously, horrifyingly slow.)
No, not exactly. He says they should outright kill the PS2.
At any rate, it's another WTF. If it still sells and makes a profit, why on Earth would any company want to kill it?
I agree with most of your points, but not this one. He's saying that the PS3 would be in a much better position if the PS2 was gone. Right now they're raking in their short-term profits (PS2 sales) and sacrificing their long-term profits (extra adoption of the PS3). He thinks that Sony is shooting themselves in the foot badly here, and they're going to regret their current actions in a year or two when the PS3 is relegated to third place even more firmly than it is right now.
Just because something is profitable now does not necessarily mean it's a good idea.
On Windows, simply opening the appropriate path opens the file - I can write a ten-line C++ program that's capable of opening "//anothercomputer/sharename/myfile.txt", without using any Windows-specific code, and many people have done exactly that without even realizing that they have. The OS takes the filename, parses it, and provides a standard file layer that just happens to work over the network. I make use of this regularly - I've actually mounted disk images across networks, for example. Works just fine. (It's even surprisingly fast.)
On Linux, all of that stuff is an absolute pain - you can't just open files off shares, you have to copy them locally, because the Linux kernel doesn't (as far as I know) yet support easy and convenient plugins of that type.
I assume OSX has taken the Windows route, which is the right solution. Linux has some catching up to do here IMHO.
Microsoft's RDP protocol, with 300kbit upstream, is quite responsive (as long as you aren't trying to display heavy graphical images). I've heard that it's better than old-style X, but there are better open-source equivalents that are being worked on.
The one problem with this is that as the protocol gets more and more complex, the horsepower needed on the clientside gets larger and larger - and eventually you end up sitting there thinking "why didn't we just put an OS on this thing in the first place".
I don't know what you're smoking, but I have a Linksys router next to me that I purchased two weeks ago. Flashed it to Tomato using the "update firmware" tab on the admin interface and nothing more - no holding down reset, no TFTP, just upload and go.
Your point (1) is simply incorrect - there's no reason whatsoever that the player couldn't use a sideband database, recognizing songs by some provided metadata and storing with those as the keys in a separate file. Nobody has done this yet, but there's no reason the set of features you want requires an iPod-like database. It could all be done perfectly adequately on a classic mass storage system.
Also, while I personally don't want any dedicated music managers that move files around or change them, there's no reason you couldn't write - again - a sideband music manager that did all that processing next to your file database, rather than on top of it. I don't need (or want) a music manager to own my files, I merely need it to be aware of their existence, and this can all be done reasonably efficiently without needing to muck about with my collection.
My ISP actually does offer DHCP to assign static addresses - my package comes with no less than 8 static IPs, and if I wanted to I could use DHCP for them.
In reality, I have an OpenBSD box acting as a router with static IPs (and NAT, natch), and on the other side I have . . . static IPs distributed over DHCP because it's just plain easier. My computers have static IPs, friend's laptops get assigned dynamic IPs, everything works.
DHCP is awesome whether you have static or dynamic IPs.
Slashdot Mirror
People seem to think that I'm planning on taking their hard work and running away with it. I'm not. The free version and the commercial version are the same toolkit. If someone contributed to my GPL project, I wouldn't be able to close it later. I don't want to steal anyone's hard-earned work. I just want the option to switch to the commercial library later.
It's a game editor. It doesn't have much value outside the game. It does have value with the game, and if I were to release it it'd be as a free addon - however, since it uses a good chunk of the game code internally, and I don't want to open-source the game, I don't want to be forced to open-source the editor also. Therefore a GPL library isn't practical.
If they were GPL-only, I'd say, sure, our philosophies differ, and I'd move on. But they aren't. They're GPL, and commercial - but for some reason they demand I pay now if there is ever a possibility that I would want to use their commercial package. That's weird.
"Just do it, there's no way they could know" is not an acceptable option, IMHO. If they wanted to avoid the many-developers one-license thing, they could simply require one license for each person who had worked on the codebase before the conversion. I'd be fine with that. But they didn't.
As it is, I'm using wxWidgets instead. wxWidgets is basically the LGPL with an added exception, which makes it both a free software library and entirely practical for proprietary software (even more so than LGPL, in fact.)
Free software and commercial software aren't mutually exclusive, and QT is neatly killing an entire segment of the market for themselves (namely, people who later decide they want to close their software). Their choice, but I have to say it's a weird decision.
I think you're missing my point somewhat - I can't, as a small developer who doesn't even know if his software is going to be released commercially, start coding now and then purchase a license later. I'm a small game developer and my editor may be of no interest to anyone but me. But if it does turn out to be useful to release it, and I don't want to release it open-source, I can't simply buy a commercial license and be done with it.
Why should Trolltech mind if I bought a license later rather than sooner? They're still getting the license. One way just forces me to decide much earlier, when I may simply not have the information that I need to determine which is the right course of action. (Which, in this case, turns out to be "don't use QT".)
QT has two different licenses - their open-source license, and their commercial license. This is not a problem - I'm fine with this.
The problem I have is that they require that any software written for their commercial-license library be only written for their commercial-license library. This means that if, like me, you're someone trying to start a game studio looking for a basic windowing library for an editor, you have three basic choices:
* Write your editor with their free library, then never be able to distribute it in any way without GPL'ing it
* Shell out $$$ for the commercial library, whether or not you'll ever need to distribute it
* Use a different library
Obviously I've chosen #3, but I can't help but think that perhaps Trolltech lost a sale there - I probably would have used QT if it had been a viable option, and if I'd ever decided to distribute the editor I likely would have gladly paid the licensing fee. It's a bizarre licensing decision.
Just click here, then turn your screen off.
Cromulent. I swear, kids these days can't spell properly.
And it works, you fall for it too. How else do you know it was a SHELL gas station? If you were imune to it and not a sheep you would just tank at any gas station. (but without any advertising whatsoever, how would you know it is a gas station?) You obviously saw Shells adversting, yes even the sign that says Shell is part of advertising.
I did the same thing that the GP did, and the only reason I know it was a Shell station is because I explicitly checked once the ads started so I'd know which gas stations to avoid in the future. I wouldn't have known it was Shell if they hadn't made me care.
Except I was, actually, willing to pay. I donate pretty frequently to open-source projects. And it may have been hard the first time they did it, but considering that they were already providing a crippled win32 version I kind of doubt the "registered" version was any harder than clicking the "make it not suck" checkbox.
I suppose it's possible that they would only have ported it to Windows if people were willing to pay - but considering that there was a third party perfectly willing to make a Windows version on its own, I don't actually believe that either.
X-Chat does this.
X-Chat is a very good IRC program. I use it on Windows. I was looking for an open-source chat program, I found X-Chat, and decided I liked it.
I didn't like their philosophy. They claimed that making a Windows build was "hard" somehow, but anyone who knows anything about programming knows that once you have the dev environment set up properly, making a new build is a matter of minutes. They'd obviously set up the dev environment properly, so why were they charging money for the Windows version?
However, there's somebody who makes Windows builds of X-Chat. For free. Unaffiliated with the actual group, of course, but fully functional. So I downloaded from him . . . and donated $10 to him for providing a useful service without demanding money.
Still haven't given a cent to the "main team" and I don't plan to until they get rid of that braindamaged policy.
If you split your userbase among bizarre lines, and demand money from one group, you may very well piss off the people who you're demanding money from. Be careful about things like this.
I'm self-employed.
A month and a half ago my fiancee dumped me. I said, essentially, "Fuck! That sucks. Okay I'm going to slack for a while." It's been a month and a half since, and I've beaten a fuckton of video games. I'm feeling a lot better now. I think once I'm done with Super Paper Mario and the Sam and Max series I'll be ready to go back to work.
I've got enough money that I can do this. I've got enough skills that I have no worries about my ability to find a new job. I live in a 900-square-foot apartment four blocks from one of the Bay Area downtown nodes, and I have a 60" TV, no cable, all the game consoles I want, and two cars, the newest of which is an '88 Sentra without a working stereo.
I'm living my life how I want to. I have the things I care for, and I don't spend money on status symbols. I keep enough reserve that I can take a break when I need it.
I've got the "something", and I can tell you, quite honestly, "something" can be totally awesome.
Now I'm gonna go have a beer and finish World 7.
"Grammar".
I've never had my DS, PSP, XBox 360, PS2, Wii, or Gamecube crash. Ever. And I've used each one of those systems more than I've ever used a graphing calculator . . . and curiously enough, I have had my TI-81 crash. Lost all its saved data too.
Modern technology doesn't imply frequent crashing. Modern technology and complex code doesn't even imply frequent crashing. I have no doubt that you could build a cheap graphing calculator on a 400mhz XScale chip, with maybe 128mb of storage, with a full GUI and a hierarchical filesystem, just as stable as they used to be.
Probably based on Linux.
I would agree with your sentiments, but many Internet Service Providers are putting in capacity limits on how much you can download per month, and that will kibosh any idea of a VC-1 or AVC encoded 1080p 48 fields/second movie.
Yes, this will be an issue. I remember some of the early ISPs with 50mb limits, and those limits really haven't been raised at all since.
Honestly, I don't have that fast of a connection, and it takes me about half an hour to download an entire CD. 600 megabytes. That 200-meg file? 10 minutes. You are living in the past, and not realizing that connections will keep getting faster.
It's highly doubtful that he'd ever accept v3-only submissions since that would mean that the kernel could no longer be distributed under gplv2 *or* gplv3, but instead different parts of it would have to be distributed under different licenses. I'm not even sure that's legally possible.
I imagine he'd accept dual-licensed submissions but it's quite possible that he wouldn't bother to document the gplv3 part.
It's not over until Netcraft confirms it!
Good point, I'd forgotten about that - some people at my high school computer lab used to run SimAnt over the network (which was, obviously, horrifyingly slow.)
Oh, I never said it was simple :) Just that there are, sometimes, reasons not to sell products, even if you make money off them.
I agree with most of your points, but not this one. He's saying that the PS3 would be in a much better position if the PS2 was gone. Right now they're raking in their short-term profits (PS2 sales) and sacrificing their long-term profits (extra adoption of the PS3). He thinks that Sony is shooting themselves in the foot badly here, and they're going to regret their current actions in a year or two when the PS3 is relegated to third place even more firmly than it is right now.
Just because something is profitable now does not necessarily mean it's a good idea.
Whenever I need to make a comment about the next generation of consoles I refer to them as the XBox 720,000, the PlayStation eXtreme, and the Wiiwii.
It shouldn't.
On Windows, simply opening the appropriate path opens the file - I can write a ten-line C++ program that's capable of opening "//anothercomputer/sharename/myfile.txt", without using any Windows-specific code, and many people have done exactly that without even realizing that they have. The OS takes the filename, parses it, and provides a standard file layer that just happens to work over the network. I make use of this regularly - I've actually mounted disk images across networks, for example. Works just fine. (It's even surprisingly fast.)
On Linux, all of that stuff is an absolute pain - you can't just open files off shares, you have to copy them locally, because the Linux kernel doesn't (as far as I know) yet support easy and convenient plugins of that type.
I assume OSX has taken the Windows route, which is the right solution. Linux has some catching up to do here IMHO.
Yeah, you've pretty much just described the system I was thinking of actually :P
Microsoft's RDP protocol, with 300kbit upstream, is quite responsive (as long as you aren't trying to display heavy graphical images). I've heard that it's better than old-style X, but there are better open-source equivalents that are being worked on.
The one problem with this is that as the protocol gets more and more complex, the horsepower needed on the clientside gets larger and larger - and eventually you end up sitting there thinking "why didn't we just put an OS on this thing in the first place".
I don't know what you're smoking, but I have a Linksys router next to me that I purchased two weeks ago. Flashed it to Tomato using the "update firmware" tab on the admin interface and nothing more - no holding down reset, no TFTP, just upload and go.
Your point (1) is simply incorrect - there's no reason whatsoever that the player couldn't use a sideband database, recognizing songs by some provided metadata and storing with those as the keys in a separate file. Nobody has done this yet, but there's no reason the set of features you want requires an iPod-like database. It could all be done perfectly adequately on a classic mass storage system.
Also, while I personally don't want any dedicated music managers that move files around or change them, there's no reason you couldn't write - again - a sideband music manager that did all that processing next to your file database, rather than on top of it. I don't need (or want) a music manager to own my files, I merely need it to be aware of their existence, and this can all be done reasonably efficiently without needing to muck about with my collection.
My ISP actually does offer DHCP to assign static addresses - my package comes with no less than 8 static IPs, and if I wanted to I could use DHCP for them.
In reality, I have an OpenBSD box acting as a router with static IPs (and NAT, natch), and on the other side I have . . . static IPs distributed over DHCP because it's just plain easier. My computers have static IPs, friend's laptops get assigned dynamic IPs, everything works.
DHCP is awesome whether you have static or dynamic IPs.