On 6/29/06, Chris chris.kcat@gmail.com wrote:
If you notice, Sony got into a lot of trouble over that. And the problem wasn't autorun. The problem was that the disc installed the rootkit anyway /even if the user said no/. The same exact thing would've happened if the user had to browse the CD and double-click setup.exe, or whatever the file was called. Should Wine disable running .exe files because they may install rootkits on users' machines? Of course not, because that would be couter-productive to what Wine is trying to achieve. It's the same thing with autorun. It may or may not cause problems, but it's the user's responsibility to take proper care of their machine. It's just as true in Windows as it is in Linux, or any other OS.
[skip to the last paragraph to get to the point; I thought this other stuff was necessary and now I don't, but I also don't feel like deleting it having already written it]
I'm not sure we should install programs in that way on Linux.
I think one of the most noticeable differences between Linux and Windows for most people is how you install software.
In Windows, you get an exe file from somewhere (hopefully from someone you can trust), and you run it. You then follow the instructions, possibly reboot the computer, and then when you're done you expect the program to be installed and working. If it doesn't work, you complain to the authors of the software.
In Linux, most of the software you need will be provided by your distribution. If you're on, say, gentoo, you type "emerge software", and when it's done you expect the program will be installed and working (in this case, you don't have to do anything; you just wait). If it doesn't work, you don't complain to the authors, you complain to gentoo. If it turns out to be an upstream bug, it gets passed along to the authors. If it turns out to be a bug in gentoo, then hopefully they fix it.
Of course, Linux users can also do something similar to what the Windows users do and handle the whole process themselves. Then they'll report problems directly to the authors.
Now, here's the part where I reward you for reading this far and tell you what this has to do with Wine. Wine users get software from the authors, try to install it themselves, and expect it to work. Then when it doesn't work they complain to Wine. That's backwards. Unless they checked the appdb first, probably no one from Wine claimed it would work to begin with. Yes, most people aren't like that, and yes, people should report bugs. But they shouldn't act like someone said it would probably work if no one has.
I think this is a good reason to have something like Wine Doors. For applications that are known to work, possibly with some override or hack, someone could write an installer and vouch for the workingness of at least some apps. People who don't know how to do things like overrides can use the installer. If something doesn't work (and you should expect it to work if there's an installer that's supposed to work on that wine version) then those users have somewhere to complain to. Conversely, applications without installers would be hit-and-miss. You'll have to do a bit more work for those, and you may be paving the way for others by submitting testing data, writing howtos, bug reports, an installer, maybe even fixing something in wine. So it's clear that they might not work.
Now, when you make someone's environment say "This CD has a program on it, shall I run it?", you're making the claim that this will probably work. That's just not true. It probably will not work as well as it's supposed to. I'd much rather have people go to a package manager, the applications database, the documentation, ANYWHERE where they will not be promised something that isn't true, even if it makes Wine seem harder to use.