Open source developers would love to do just that. But if you're trying to play a DVD under Linux, you'll run into problems: To play the DVD, you need to decrypt it. This is impossible to do legally with an open source application, as it violates the American Digital Millennium Copyright Act, which prohibits circumvention of copy prevention mechanisms. (Even creating a legal, closed-source player requires payment of licensing fees for the keys, which is impossible for a Linux distribution that can be downloaded and distributed freely.) There may eventually be a commercial closed source player, but that is obviously incompatible with the whole open source idea.
The DMCA will be implemented in different variants world-wide. This is a real issue: To play DVDs on Linux, you need to break the law, in America and soon elsewhere as well. That's why it's important to change the law instead of just passively ignoring what's going on and hoping that the problems will go away. If you do that, what's currently the case with DVDs will soon be the case with all commercial media, thereby defeating the whole point of open source.
Note that the copyright cartels have already successfully gone after people who distributed the DVD decryption software, and even those who linked to the tool that allows doing so. They love the additional control over content use that the DMCA gives them, and they'll fight to keep it and to extend it even further (which brings us to Microsoft's Palladium).
On the other hand, the GDF has nothing to do with the original, discontinued Gnutella client by Justin Frankel. It was started separately by some, but not all Gnutella client developers of the time. So I don't really see how they have any right whatsoever to the Gnutella name.
Re:After my last month with RH8 M$ has no worries
on
Halloween VII
·
· Score: 2
The Linux HOWTOs and man pages are maintained by volunteers, not by the distributors. There would be ways to keep them more up to date, such as use of wikis (collaboratively editable webpages), but this is just beginning to happen (see, for example, LinuxWiki).
As a Linux user, you can expect the explicitly listed features of your distribution to work. If you are sufficiently technical, you can hunt down the necessary information to make extra features work, but you should realize that you're now on your own, and working with information provided to you free of charge. Instead of respecting those who put effort into doing so (whereas it would be much easier for them to just keep the info for themselves, as you likely will once you make things work), you criticize them. At least send them a mail and tell them where the error is.
You should realize one thing: Red Hat 8 is not geared towards people who burn VCDs. It's geared towards corporations that want to replace their Unix/Windows servers and workstations. For them, it does a very good job. The corporations save licensing costs, and Red Hat earns money with support. Home users are, from the perspective of Red Hat (the company) just evangelists, small change or even parasites. Thus, your expectations towards Red Hat to be a good solution for a home desktop are misguided. If you're looking for that, try Xandros (which is, however, also geared primarily towards corporate use) or Lindows.
Only a year ago, every Slashdot story about Linux on the desktop would be about how Linux would never succeed as a desktop platform. Now people are complaining about small issues. Are you beginning to see a pattern emerge?
Re:After my last month with RH8 M$ has no worries
on
Halloween VII
·
· Score: 1
Uh, now you're making a completely different statement than the original poster, namely that Linux is "still years behind on the interface and ease of use level". Back that up, please.
Re:After my last month with RH8 M$ has no worries
on
Halloween VII
·
· Score: 2
Some enlightenment:
1) nVidia's drivers cannot be bundled by Red Hat for licensing reasons. If you want them to be bundled and auto-installed, petition nVidia to release their drivers as free software.
2) New hardware isn't exactly an advantage when using Linux, because support for it is often provided by volunteers and needs time to get mature. If you want official Linux drivers, petition the manufacturer. This is not a problem of Linux, it's a problem of any OS that wants to compete with a monopoly OS. A Linux-preinstalled machine does obviously not have this problem, but MS has so far prevented dual boot Windows/Linux machine sales through OEM pressure.
3) Users who want to install the latest software from source or CVS should expect this to be a non-trivial procedure, regardless of the OS being Linux or Windows (the latter of which comes with QBASIC and VBScript as its only development tools, the better one, QBASIC, is no longer part of recent releases). If you want to install software that immediately runs, use your distribution's packaging and installation system. In the Red Hat case, this is called up2date and is commercial. You can also spend time instead of money and install something like apt4rpm to make package installation simpler. You can also use Ximian's Red Carpet for free (but it is primarily geared towards GNOME applications). Other distributions like Debian (which you can try out using the fantastic Knoppix) provide even more sophisticated mechanisms.
4) I have used OpenOffice Impress. It's somewhat unstable but imports PowerPoint presentations reasonably well and has most of PowerPoint's features (and some of its own). I have found the script-based MagicPoint more satisfying in getting quick and pretty results (example pres I did about Mono). If you get over the fact that it isn't yet another PowerPoint clone but actually a different way to do things, it's pretty cool.
KPresenter may eventually bcome the best graphical presentation tool, but is not there yet. You can run PowerPoint nicely under Linux using Crossover Office and, probably, with some tweaking, under the free WINE.
5) Homemade karaoke VCDs: Exotic end user stuff like that usually takes extra effort on Linux because too few people care about it to develop free, easy to use apps, and desktop Linux is not yet sufficiently wide-spread to be commercially targeted for such applications. Obviously, the best way to change this is to stop using the monopoly OS and to use Linux instead, or to fund development efforts.
6) 3D speed: I don't play FPS, so I can't comment much on that. Last time I tried 3D stuff under Linux, it worked as intended, so I didn't check the FPS. DirectX is obviously a quite sophisticated API and the Win32 drivers are highly optimized, though, so until Linux game companies start targeting Linux as a major platform, I wouldn't be surprised by about 10-20% speed differences.
In conclusion, your problems resulted from you doing stuff that basic users shouldn't do unless willing to spend the effort (trying to install software from source), not using one of the free or commercial software installation tools, and not checking hardware compatibility properly. Many of the problems are not problems of Linux as an OS but problems of a market dominated by a monopolist. Therefore, your attitude that you hope that Linux will "fix" these problems is somewhat naive, the way to fix them is to support Linux so there is no longer a monopolist who imposes market conditions under which much of the software you miss so dearly (drivers, Karaoke, games) can be developer.
Obviously, you can also continue to use Windows. It's sufficient for basic desktop stuff, and Microsoft has some very interesting and powerful features coming up, like "trusted computing". But while you continue to buy their software, you are part of the problems you criticize.
I also recently discovered MyCC (which is open source) and I agree, it's a great new front-end that simplifies usage a lot. I used MySQL-Navigator before, which isn't bad but nowhere nearly as feature rich as MyCC. MyCC has both an SDI and an MDI interface and lets you directly edit data in the table display. Since it uses a Qt-based interface the GUI should immediately feel familiar to Windows users.
Take note that MyCC is still alpha software, though, so crashes are to be expected. I'm glad that MySQL AB is working on this, a nice graphical client should really speed up the adoption of open source databases. Now an Access-style desktop database with a form designer would be nice -- the GNU Enterprise project is working on this. theKompany also has a proprietary product called Rekall. Any others?
On a system with one user/administrator, it's meaningless.
Bullshit. My system is set up in such a way that all data I no longer work on frequently is moved into a non-writable archive. That also includes all media files like MP3s, movies etc. If a malicious executable should wreck havoc, it can delete a few days of work at most. (Backups are nice, but even nicer if you don't need them.) I see no reason why this shouldn't be wrapped into a nice GUI and made standard behavior.
As for possible damage, SuSE's example for the damage root can cause has always been a user who typed "ls >/dev/hdb1" (he wanted to redirect it to the sound device to see what would happen). Similar accidents can happen easily with drag & drop, and Unix is a lot better at deleting files that are in use than Windows. Trust tech support wisdom: users do stupid things. It's irresponsible to put them in an environment where they can cause a lot of harm, especially when we're trying so hard to teach newbies that they can experiment freely without being afraid of computers. If you're root, better be afraid.
You're right that non-root users can still cause damage to others (although they can't listen on low ports, for example), but the infection is much easier to fix. Getting rid of a kernel-based rootkit and cleaning infected system binaries is hard, removing something from.xinitrc or.kde/Autostart is not. It's really not that difficult: There's a difference in having write access to 2 megs of files versus the entire system.
Lindows is dangerous and should be boycotted by all security-conscious users. The reason is simple: Users run as "root" by default, with all rights -- a single wrong click or command and the whole system is made unusable. Or turned into a full-powered skript kiddie battle station.
This kind of philosophy has been the main cause of many destructive worms and viruses on the Windows platform. To repeat this error endangers the Internet ecosystem as a whole and gives Linux a bad name. Furthermore, it gives people a justification to run as root -- this practice should be discouraged. Any operating system that is insecure by default should be boycotted.
Lindows.com is currently stating that they are doing this in the name of convenience, a stupid argument (how hard can it be to ask for an administration password?). As long as they do not reverse their stance in this matter, Lindows should be boycotted by all technically competent users. I'm getting enough e-mail worms per day as it is.
1) What's your relationship with Michael Robertson, given that you are working together on the Sincere Choice project?
2) Have you talked to him about the "always running as root" issue? That's a big security risk that puts Lindows on the same level as Windows 95.
Giving away StarOffice != giving away Windows
on
Sun To Sell Linux PCs
·
· Score: 5, Informative
First, you should know that the author of the editorial is the same troll that wrote brilliant articles like "The jihad against Microsoft":
Torvalds posted his Unix rip-off dubbed "Linux" on the Internet in 1991 for free. True to his family's socialist radical politics, Torvalds released his OS under the non-standard General Public License (GPL) or "copyleft." Under the GPL, programmers had the ability to download Torvalds' Linux, fix the bugs in his program and give the improved program back to him to distribute to the Linux community. GPL programs are essentially community property with no real owners, but since Torvalds was the originator of the rip-off, it becomes his personal rip-off to control as he wishes. In other words, Torvalds became the dictatorial leader of the Linux cult with all decisions for the greater community good going through him first, then doled out at his convenience.
Let's all move Scott McCollum into our collective killfiles and move on, shall we? Furthermore, the key difference between Sun's donation and Microsoft's, besides the fact that Sun is not a monopoly, is that Sun has open sourced Star Office. To gloss over this little fact is typical for a professional troll like McCollum. While Star Office itself is not open, it's an open platform, and the differences between SO and OO are minor. So even if SO/OO were to become the standard, it would always be easy to move somewhere else if necessary (and you can bet someone will fork OO if Sun does something fishy).
You may not have noticed it, but sites differ in their target audience. I would very much doubt that only 1% of Slashdot's or K5's visitors use Mozilla, for example.
So be careful the next time you visit that open source news site..
I don't know about you, but I started using MP3 when it first came out, and my only machine back then was a 386 or 486. I had no problems (my father did though, stereo playback would overtax his machine), but I can't give you the exact specs of the PC I used. I would imagine that current decoders can do even better, but I'd like to hear from some people who have tested it (with mp3blaster for Linux, for example).
1) You may not have noticed it, but not all computers come with Windows licenses, although Microsoft does everything they can to make sure that this changes. Buying a new machine without Windows can easily save you $100, and used machines are often blanked before they are sold.
2) What applications are you going to run if you get a cheap machine with a Windows license? Microsoft Paint and WordPad? Linux comes with thousands of free, powerful apps, many of which run on low cost hardware. Besides the fact that you will have to pay for them, apps that can be bought today will typically not run on low cost hardware, and older apps are often deliberately taken off the market. (Piracy is obviously an option, but in the long term only increases the dependence on a software oligarchy.)
3) If you decide to use a cheap Windows (95/98) license anyway, you're stuck with an unsupported operating system that's still based on DOS, horribly unstable, wide open security-wise, and that will neither work with future hardware nor future software (regular forced upgrade cycles are necessary to keep the OS market running, you know).
Aside from that, even the claim that Win98 will run faster than a light X configuration is debatable (I actually compared both when a P166 was damn fast -- applications under X would typically take longer to load, but work faster and multitask better once loaded). Certainly, recent scaled down versions of Linux for embedded devices will give Win98 a run for its money.
Desktop usage != web usage. US web usage makes up the largest share by far of international web usage: 42.65%, followed with considerable distance by China (6.63%). Since Microsoft is ultra-dominant in the US, this skews the data. A lot of threshold nations have a large amount of PCs but relatively little Internet use, mostly for cost reasons. And let's not forget censorship -- China recently censored Google, for example.
One great advantage of Linux, besides being free, is that when correctly tuned it works on very cheap hardware. Even if you just have a 386 or 486, you can still use thousands of decent console applications (including stuff like MP3 players and web browsers -- heck, you can even use mplayer with an EGA graphics card) and get drivers for modern hardware. An old Pentium is fast enough for a simple X11 setup with small desktop aps like WindowMaker, LyX etc.
That being said, I don't buy the 3.9% number without some supporting evidence. Even in developing nations Windows is only slowly being replaced by Linux desktops, with relatively few major rollouts in recent months, and while Linux can run on low cost computers, the problem is that it's not exactly easy as pie to tune and configure properly. Internationalization is another issue..
This is the list of sites that have been found to be inaccsesible. A lot of them are the expectable human rights (Amnesty etc.), Tibet and Falun Gong stuff, as well as some news media (Yahoo Asia News, CBS News, BBC news, and many US-based China news sites).
Geocities appears to be completely blocked.
The Chinese government doesn't like Playboy or sex.com - hmm, do we see a correlation between repressive government and antisexual morals there? Nah, couldn't be.
Google is on their shitlist. No surprise given its cache and large index. The Wayback Machine isn't - I'd expect that to change in the long term.
Anonymizer is accessible as well.
Peek-A-Booty and Freenet are not accessible, of course. It appears that all SourceForge sites are blocked (unless the testing engine is slashdotted and not working properly, but other sites are reported as accessible). I presume this might be because Freenet is hosted at SourceForge.
I don't know their application so I can't say for sure, but in most cases that's ass-backwards. You usually want to build your product for the biggest market first.
And that's exactly what feeds Microsoft's monopoly. I applaud everyone who has the courage to try producing commercial software, even proprietary one, for the Linux market.
I think you're touching on the same two major problems that are being touched on again and again by all people who argue that Linux is not ready for the desktop -- and I agree. The problems are X' font handling (and the complexity of X11 in general) and the software installation.
For the font handling I don't really see a solution on the horizon (I don't know how Mandrake does it -- but I can hardly imagine that it solves the glaring incompatibilites between how different Linux apps handle their fonts, encodings etc.). I'm not very happy with X' slowness in general (window resizing, moving etc.) and hope that it will sooner or later be replaced by something more suitable for desktop use like Berlin or DirectFB. Hopefully, this will also bring us reasonable font management, nice and fast anti-aliasing etc. Certification for apps that comply with some minimum standards would also be nice.
One key problem here is that the combined efforts for desktop Linux are relatively modest -- SuSE, Red Hat, IBM, HP, Sun, these are all interested in the server market and don't care much about the desktop. Mandrake and Lycoris are different, but Mandrake is running out of cash and Lycoris never had much cash to begin with. The one billion dollars IBM invested in Linux are almost exclusively used for pushing it into the server market.
The other major problem of LotD, software installation (dependency hell), is virtually solved with Debian and Gentoo (although a source-based distribution is obviously only useful to very few people). Mandrake is, this is my understanding, working on a similar solution to apt-get (urpmi). And apt-rpm seems to be gaining some following. Red Hat wants to make money with up2date, so they're not interested in a free udpate service.
I'd love for the people to agree on a standard -- a lot of work is being duplicated here, resources wasted unnecessarily -- but distributed groups rarely agree on standards, because of competition and personal vanity. This is not a problem unique to open source, just look at the DVD mess. In fact, OSS might have better chances to solve these problems if some distributors (Red Hat: "We don't need the LSB, problems will solve themselves" -- Debian: "We don't need to agree on a packaging standard! Ours is technically superior, it should be used by everyone!") weren't so bone-headed.
One problem you don't touch upon is the X clipboard, which, while it's making progress, is still very much inferior to the almost universal Windows clipboard, not to mention extremely buggy (have had huge problems pasting large amounts of text into different browsers, for example). But this is related to X11's general problems and complexity -- X11 is really slowing Linux progress in my opinion.
You make a mistake if you believe that Linux users "like" these problems -- they have just learned to deal with them. But nobody with any amount of sanity left could argue that X' font handling is anything but modularity gone horribly wrong. Only the unfortunately very loud zealots seem to believe that if your denial is strong enough, the problems will magically disappear. In this respect, MS could learn a lot from OSS: If you create a community of true believers, they will do everything to cover up the flaws of your product.
He's actually right about the soft (visual) wrapping editor issue. Traditionally, Un*x editors insert hard line breaks, which is convenient for email and to some degree for programming, but hardly for anything else. What you usually want is to wrap words (not characters) at the window's edge, without inserting the breaks into the file. Even emacs doesn't seem to do it properly (there is a longlines elisp script, which is far from perfect).
I've found two notable exceptions:
vim supports visual wrapping (":set lbr") quite well. Of course, vim is a nightmare from a usability perspective, but there is a young project called Cream, which is a set of configuration files for the graphical version of vim (gvim) that turns gvim into a modeless editor behaving in many ways like typical Windows editors, while retaining vim's functionality.
I currently use gvim+cream for all my editing and am rather happy with it.
Nedit is the other exception, but I found it unusable -- hotkeys wouldn't work, dialog boxes would have six times their normal size etc. - probably some X configuration stuff, but I don't have time for fixing this. If it works on your system, it may well be a good standard editor.
None of the KDE editors in the versions I have tried supported visual wrapping, nor did any of the GNOME editors (gnp does do it, but it's extremely buggy -- when you hit "cursor down", it jumps to the next paragraph instead of the next line, which is unacceptable). Unlike some other poster claimed, gedit, at least in my version (.96), doesn't do visual wrapping. For KDE's showcase editor Kate it's apparently in the works.
Yes, you can use something like abiword, but honestly, abiword is generally a PITA and uses the ugliest screen fonts in the known universe, and who would want to start OpenOffice for editing a text file? Generally, I consider the lack of a properly behaving, usable text editor a big problem and would contribute financially to any project aiming to fix this.
Thanks to all the developers for making it happen, and thanks to AOL for funding an open source effort. Mozilla has been my only browser for the last few months and I'm unlikely to switch again (unless Opera becomes open source *cough*).
You can find some cool add-ons for Mozilla at Mozdev. Among these are: Annozilla, a sidebar tool for annotating websites; Forumzilla, a tool for reading web discussion forums usenet-style; Jabberzilla, a Jabber-client; MozBlog for weblog authors; OptiMoz for mouse gestures, and many many others. Not all of these work with 1.0 yet, though.
You should read before you respond. I did not claim XP has a "command line process to its setup". Do you even know the difference between a command line and a text mode application? Most Linux installers are fully graphical and have Windows-like look & feel from the start, the Windows install isn't. And you did not get any of my other points with regard to "simplicity" either.
You're probably talking about the last phase of the install that is run on the user's deployed machine. The actual Windows XP installer is based on NT's and text-menu based. And as I explained in my other comment, it is far inferior in functionality to all Linux installers (with the exception of hardware detection).
Unless Linux's setup has gotten astronomically better since Red Hat 5.0
It has. Try Lycoris or Mandrake. Nowadays you play Solitaire during the install. Of course not all hardware is supported, but that, again, is the result of Microsoft's monopoly.
MS's setup is the best OS setup I have ever used, period.
Then why doesn't the Windows XP installer recognize my FreeBSD and Linux partitions and allow me to select them from its boot manager, or allow me to resize or create any non-Windows file system? That's right, because Microsoft has a monopoly and doesn't need to implement certain functionality others do need to implement. Feature-wise, Linux installers are far superior.
The problem is that Linux cannot gain ground until PC makers can ship dual boot systems without being punished by Microsoft. People are not going to buy a Linux-only system, they want to try Linux, but also have a safe recourse in case it doesn't work for them. The OEM channel is Microsoft's strongest defense against any competing operating system.
You may say "But what if we make installation so easy that people can just do it later?" That's a flawed premise. Installation of Linux is already fairly easy, especially when compared with Windows' primitive text-based installer that can hardly do anything. Besides the fact that most people are never going to bother with the installation of a new OS, the problem is that people who convert to Linux will want to preserve their existing Windows systems. To do so, they will have to resize their existing partitions, which are increasingly in Microsoft's proprietary NTFS file system format. Resizing NTFS partitions, to my knowledge, is not possible with any Linux installer, and if it is made possible, MS can threaten to sue those who implement it over their NTFS patents (as they have done in the past), as well as alter the standard unpredictably. This makes it almost impossible to implement simple dual boot installation, unless you're willing to piggyback on NTFS and the Windows bootloader -- generally a bad idea for obvious reasons.
Simply put, if Microsoft keeps the OEM channel, gaining ground outside schools and developing countries will be hard.
I read Voodo Science. It's a good book and gives a nice summary of subjects like homeopathy and manned space exploration. What it lacks the most are sources. The author states that he didn't want his book to be riddled with footnotes so as not to confuse the reader, but that is obviously a stupid attitude for a book that is written to encourage people to embrace science. Author Robert Park also writes a newsletter called What's New about developments in Voodo Science.
Park's book should be read together with another one: Trust Us, We're Experts! (Amazon) by John Stauber and Sheldon Rampton. While there is a lot of "junk science" out there, there is at least as much corporate sponsorship behind efforts to discredit real scientific work as such. See also this story about PR efforts to discredit global warming, and my related K5 comment.
Slashdot Mirror
The DMCA will be implemented in different variants world-wide. This is a real issue: To play DVDs on Linux, you need to break the law, in America and soon elsewhere as well. That's why it's important to change the law instead of just passively ignoring what's going on and hoping that the problems will go away. If you do that, what's currently the case with DVDs will soon be the case with all commercial media, thereby defeating the whole point of open source.
Note that the copyright cartels have already successfully gone after people who distributed the DVD decryption software, and even those who linked to the tool that allows doing so. They love the additional control over content use that the DMCA gives them, and they'll fight to keep it and to extend it even further (which brings us to Microsoft's Palladium).
On the other hand, the GDF has nothing to do with the original, discontinued Gnutella client by Justin Frankel. It was started separately by some, but not all Gnutella client developers of the time. So I don't really see how they have any right whatsoever to the Gnutella name.
As a Linux user, you can expect the explicitly listed features of your distribution to work. If you are sufficiently technical, you can hunt down the necessary information to make extra features work, but you should realize that you're now on your own, and working with information provided to you free of charge. Instead of respecting those who put effort into doing so (whereas it would be much easier for them to just keep the info for themselves, as you likely will once you make things work), you criticize them. At least send them a mail and tell them where the error is.
You should realize one thing: Red Hat 8 is not geared towards people who burn VCDs. It's geared towards corporations that want to replace their Unix/Windows servers and workstations. For them, it does a very good job. The corporations save licensing costs, and Red Hat earns money with support. Home users are, from the perspective of Red Hat (the company) just evangelists, small change or even parasites. Thus, your expectations towards Red Hat to be a good solution for a home desktop are misguided. If you're looking for that, try Xandros (which is, however, also geared primarily towards corporate use) or Lindows.
Only a year ago, every Slashdot story about Linux on the desktop would be about how Linux would never succeed as a desktop platform. Now people are complaining about small issues. Are you beginning to see a pattern emerge?
Uh, now you're making a completely different statement than the original poster, namely that Linux is "still years behind on the interface and ease of use level". Back that up, please.
1) nVidia's drivers cannot be bundled by Red Hat for licensing reasons. If you want them to be bundled and auto-installed, petition nVidia to release their drivers as free software.
2) New hardware isn't exactly an advantage when using Linux, because support for it is often provided by volunteers and needs time to get mature. If you want official Linux drivers, petition the manufacturer. This is not a problem of Linux, it's a problem of any OS that wants to compete with a monopoly OS. A Linux-preinstalled machine does obviously not have this problem, but MS has so far prevented dual boot Windows/Linux machine sales through OEM pressure.
3) Users who want to install the latest software from source or CVS should expect this to be a non-trivial procedure, regardless of the OS being Linux or Windows (the latter of which comes with QBASIC and VBScript as its only development tools, the better one, QBASIC, is no longer part of recent releases). If you want to install software that immediately runs, use your distribution's packaging and installation system. In the Red Hat case, this is called up2date and is commercial. You can also spend time instead of money and install something like apt4rpm to make package installation simpler. You can also use Ximian's Red Carpet for free (but it is primarily geared towards GNOME applications). Other distributions like Debian (which you can try out using the fantastic Knoppix) provide even more sophisticated mechanisms.
4) I have used OpenOffice Impress. It's somewhat unstable but imports PowerPoint presentations reasonably well and has most of PowerPoint's features (and some of its own). I have found the script-based MagicPoint more satisfying in getting quick and pretty results (example pres I did about Mono). If you get over the fact that it isn't yet another PowerPoint clone but actually a different way to do things, it's pretty cool.
KPresenter may eventually bcome the best graphical presentation tool, but is not there yet. You can run PowerPoint nicely under Linux using Crossover Office and, probably, with some tweaking, under the free WINE.
5) Homemade karaoke VCDs: Exotic end user stuff like that usually takes extra effort on Linux because too few people care about it to develop free, easy to use apps, and desktop Linux is not yet sufficiently wide-spread to be commercially targeted for such applications. Obviously, the best way to change this is to stop using the monopoly OS and to use Linux instead, or to fund development efforts.
6) 3D speed: I don't play FPS, so I can't comment much on that. Last time I tried 3D stuff under Linux, it worked as intended, so I didn't check the FPS. DirectX is obviously a quite sophisticated API and the Win32 drivers are highly optimized, though, so until Linux game companies start targeting Linux as a major platform, I wouldn't be surprised by about 10-20% speed differences.
In conclusion, your problems resulted from you doing stuff that basic users shouldn't do unless willing to spend the effort (trying to install software from source), not using one of the free or commercial software installation tools, and not checking hardware compatibility properly. Many of the problems are not problems of Linux as an OS but problems of a market dominated by a monopolist. Therefore, your attitude that you hope that Linux will "fix" these problems is somewhat naive, the way to fix them is to support Linux so there is no longer a monopolist who imposes market conditions under which much of the software you miss so dearly (drivers, Karaoke, games) can be developer.
Obviously, you can also continue to use Windows. It's sufficient for basic desktop stuff, and Microsoft has some very interesting and powerful features coming up, like "trusted computing". But while you continue to buy their software, you are part of the problems you criticize.
Take note that MyCC is still alpha software, though, so crashes are to be expected. I'm glad that MySQL AB is working on this, a nice graphical client should really speed up the adoption of open source databases. Now an Access-style desktop database with a form designer would be nice -- the GNU Enterprise project is working on this. theKompany also has a proprietary product called Rekall. Any others?
Bullshit. My system is set up in such a way that all data I no longer work on frequently is moved into a non-writable archive. That also includes all media files like MP3s, movies etc. If a malicious executable should wreck havoc, it can delete a few days of work at most. (Backups are nice, but even nicer if you don't need them.) I see no reason why this shouldn't be wrapped into a nice GUI and made standard behavior.
As for possible damage, SuSE's example for the damage root can cause has always been a user who typed "ls > /dev/hdb1" (he wanted to redirect it to the sound device to see what would happen). Similar accidents can happen easily with drag & drop, and Unix is a lot better at deleting files that are in use than Windows. Trust tech support wisdom: users do stupid things. It's irresponsible to put them in an environment where they can cause a lot of harm, especially when we're trying so hard to teach newbies that they can experiment freely without being afraid of computers. If you're root, better be afraid.
You're right that non-root users can still cause damage to others (although they can't listen on low ports, for example), but the infection is much easier to fix. Getting rid of a kernel-based rootkit and cleaning infected system binaries is hard, removing something from .xinitrc or .kde/Autostart is not. It's really not that difficult: There's a difference in having write access to 2 megs of files versus the entire system.
This kind of philosophy has been the main cause of many destructive worms and viruses on the Windows platform. To repeat this error endangers the Internet ecosystem as a whole and gives Linux a bad name. Furthermore, it gives people a justification to run as root -- this practice should be discouraged. Any operating system that is insecure by default should be boycotted.
Lindows.com is currently stating that they are doing this in the name of convenience, a stupid argument (how hard can it be to ask for an administration password?). As long as they do not reverse their stance in this matter, Lindows should be boycotted by all technically competent users. I'm getting enough e-mail worms per day as it is.
2) Have you talked to him about the "always running as root" issue? That's a big security risk that puts Lindows on the same level as Windows 95.
Let's all move Scott McCollum into our collective killfiles and move on, shall we? Furthermore, the key difference between Sun's donation and Microsoft's, besides the fact that Sun is not a monopoly, is that Sun has open sourced Star Office. To gloss over this little fact is typical for a professional troll like McCollum. While Star Office itself is not open, it's an open platform, and the differences between SO and OO are minor. So even if SO/OO were to become the standard, it would always be easy to move somewhere else if necessary (and you can bet someone will fork OO if Sun does something fishy).
So be careful the next time you visit that open source news site ..
I don't know about you, but I started using MP3 when it first came out, and my only machine back then was a 386 or 486. I had no problems (my father did though, stereo playback would overtax his machine), but I can't give you the exact specs of the PC I used. I would imagine that current decoders can do even better, but I'd like to hear from some people who have tested it (with mp3blaster for Linux, for example).
2) What applications are you going to run if you get a cheap machine with a Windows license? Microsoft Paint and WordPad? Linux comes with thousands of free, powerful apps, many of which run on low cost hardware. Besides the fact that you will have to pay for them, apps that can be bought today will typically not run on low cost hardware, and older apps are often deliberately taken off the market. (Piracy is obviously an option, but in the long term only increases the dependence on a software oligarchy.)
3) If you decide to use a cheap Windows (95/98) license anyway, you're stuck with an unsupported operating system that's still based on DOS, horribly unstable, wide open security-wise, and that will neither work with future hardware nor future software (regular forced upgrade cycles are necessary to keep the OS market running, you know).
Aside from that, even the claim that Win98 will run faster than a light X configuration is debatable (I actually compared both when a P166 was damn fast -- applications under X would typically take longer to load, but work faster and multitask better once loaded). Certainly, recent scaled down versions of Linux for embedded devices will give Win98 a run for its money.
One great advantage of Linux, besides being free, is that when correctly tuned it works on very cheap hardware. Even if you just have a 386 or 486, you can still use thousands of decent console applications (including stuff like MP3 players and web browsers -- heck, you can even use mplayer with an EGA graphics card) and get drivers for modern hardware. An old Pentium is fast enough for a simple X11 setup with small desktop aps like WindowMaker, LyX etc.
That being said, I don't buy the 3.9% number without some supporting evidence. Even in developing nations Windows is only slowly being replaced by Linux desktops, with relatively few major rollouts in recent months, and while Linux can run on low cost computers, the problem is that it's not exactly easy as pie to tune and configure properly. Internationalization is another issue ..
Not for me - I get the inaccessibility message. Their testing method is probably unreliable.
Geocities appears to be completely blocked.
The Chinese government doesn't like Playboy or sex.com - hmm, do we see a correlation between repressive government and antisexual morals there? Nah, couldn't be.
I have no idea why they censor {Insert Something Funny}, an obscure weblog, an anti-tobacco group, the Canadian Parks and Wilderness Society, Columbia Earthscape, or Columbia University.
Google is on their shitlist. No surprise given its cache and large index. The Wayback Machine isn't - I'd expect that to change in the long term. Anonymizer is accessible as well.
Peek-A-Booty and Freenet are not accessible, of course. It appears that all SourceForge sites are blocked (unless the testing engine is slashdotted and not working properly, but other sites are reported as accessible). I presume this might be because Freenet is hosted at SourceForge.
And that's exactly what feeds Microsoft's monopoly. I applaud everyone who has the courage to try producing commercial software, even proprietary one, for the Linux market.
For the font handling I don't really see a solution on the horizon (I don't know how Mandrake does it -- but I can hardly imagine that it solves the glaring incompatibilites between how different Linux apps handle their fonts, encodings etc.). I'm not very happy with X' slowness in general (window resizing, moving etc.) and hope that it will sooner or later be replaced by something more suitable for desktop use like Berlin or DirectFB. Hopefully, this will also bring us reasonable font management, nice and fast anti-aliasing etc. Certification for apps that comply with some minimum standards would also be nice.
One key problem here is that the combined efforts for desktop Linux are relatively modest -- SuSE, Red Hat, IBM, HP, Sun, these are all interested in the server market and don't care much about the desktop. Mandrake and Lycoris are different, but Mandrake is running out of cash and Lycoris never had much cash to begin with. The one billion dollars IBM invested in Linux are almost exclusively used for pushing it into the server market.
The other major problem of LotD, software installation (dependency hell), is virtually solved with Debian and Gentoo (although a source-based distribution is obviously only useful to very few people). Mandrake is, this is my understanding, working on a similar solution to apt-get (urpmi). And apt-rpm seems to be gaining some following. Red Hat wants to make money with up2date, so they're not interested in a free udpate service.
I'd love for the people to agree on a standard -- a lot of work is being duplicated here, resources wasted unnecessarily -- but distributed groups rarely agree on standards, because of competition and personal vanity. This is not a problem unique to open source, just look at the DVD mess. In fact, OSS might have better chances to solve these problems if some distributors (Red Hat: "We don't need the LSB, problems will solve themselves" -- Debian: "We don't need to agree on a packaging standard! Ours is technically superior, it should be used by everyone!") weren't so bone-headed.
One problem you don't touch upon is the X clipboard, which, while it's making progress, is still very much inferior to the almost universal Windows clipboard, not to mention extremely buggy (have had huge problems pasting large amounts of text into different browsers, for example). But this is related to X11's general problems and complexity -- X11 is really slowing Linux progress in my opinion.
You make a mistake if you believe that Linux users "like" these problems -- they have just learned to deal with them. But nobody with any amount of sanity left could argue that X' font handling is anything but modularity gone horribly wrong. Only the unfortunately very loud zealots seem to believe that if your denial is strong enough, the problems will magically disappear. In this respect, MS could learn a lot from OSS: If you create a community of true believers, they will do everything to cover up the flaws of your product.
I've found two notable exceptions:
I currently use gvim+cream for all my editing and am rather happy with it.
None of the KDE editors in the versions I have tried supported visual wrapping, nor did any of the GNOME editors (gnp does do it, but it's extremely buggy -- when you hit "cursor down", it jumps to the next paragraph instead of the next line, which is unacceptable). Unlike some other poster claimed, gedit, at least in my version (.96), doesn't do visual wrapping. For KDE's showcase editor Kate it's apparently in the works.
Yes, you can use something like abiword, but honestly, abiword is generally a PITA and uses the ugliest screen fonts in the known universe, and who would want to start OpenOffice for editing a text file? Generally, I consider the lack of a properly behaving, usable text editor a big problem and would contribute financially to any project aiming to fix this.
You can find some cool add-ons for Mozilla at Mozdev. Among these are: Annozilla, a sidebar tool for annotating websites; Forumzilla, a tool for reading web discussion forums usenet-style; Jabberzilla, a Jabber-client; MozBlog for weblog authors; OptiMoz for mouse gestures, and many many others. Not all of these work with 1.0 yet, though.
You should read before you respond. I did not claim XP has a "command line process to its setup". Do you even know the difference between a command line and a text mode application? Most Linux installers are fully graphical and have Windows-like look & feel from the start, the Windows install isn't. And you did not get any of my other points with regard to "simplicity" either.
You're probably talking about the last phase of the install that is run on the user's deployed machine. The actual Windows XP installer is based on NT's and text-menu based. And as I explained in my other comment, it is far inferior in functionality to all Linux installers (with the exception of hardware detection).
It has. Try Lycoris or Mandrake. Nowadays you play Solitaire during the install. Of course not all hardware is supported, but that, again, is the result of Microsoft's monopoly.
MS's setup is the best OS setup I have ever used, period.
Then why doesn't the Windows XP installer recognize my FreeBSD and Linux partitions and allow me to select them from its boot manager, or allow me to resize or create any non-Windows file system? That's right, because Microsoft has a monopoly and doesn't need to implement certain functionality others do need to implement. Feature-wise, Linux installers are far superior.
You may say "But what if we make installation so easy that people can just do it later?" That's a flawed premise. Installation of Linux is already fairly easy, especially when compared with Windows' primitive text-based installer that can hardly do anything. Besides the fact that most people are never going to bother with the installation of a new OS, the problem is that people who convert to Linux will want to preserve their existing Windows systems. To do so, they will have to resize their existing partitions, which are increasingly in Microsoft's proprietary NTFS file system format. Resizing NTFS partitions, to my knowledge, is not possible with any Linux installer, and if it is made possible, MS can threaten to sue those who implement it over their NTFS patents (as they have done in the past), as well as alter the standard unpredictably. This makes it almost impossible to implement simple dual boot installation, unless you're willing to piggyback on NTFS and the Windows bootloader -- generally a bad idea for obvious reasons.
Simply put, if Microsoft keeps the OEM channel, gaining ground outside schools and developing countries will be hard.
Park's book should be read together with another one: Trust Us, We're Experts! (Amazon) by John Stauber and Sheldon Rampton. While there is a lot of "junk science" out there, there is at least as much corporate sponsorship behind efforts to discredit real scientific work as such. See also this story about PR efforts to discredit global warming, and my related K5 comment.