I tend not to go to the sort of sites that do this, but I'd look for "promo" sites, the kind of thing that push a product and tend to have lots of annoying animation and music. The sort of site that's generally a pain in the ass to use.
At one point (not still the case) the official Star Trek site complained about anything other than IE. Sites for new movies are another good bet.
Apple has built themselves a very profitable niche, and is happy with it. Yes, they could get more market share (but much less money, at least in the short run) by porting Mac OS X to x86. They've chosen not to. They want to maintain a market with users who are comfortable paying high prices for a polished black box system. Nothing wrong with that, but Apple is not aiming at the masses, which buy computers based on price, where Apple simply is not competitive (and, again, has chosen not to be).
Aside from the folks that use Linux on PowerPC, Apple's Mac OS doesn't really even compete with Linux all that much. People using Linux on the desktop are generally on x86 -- a new Linux user means a vanished Windows users.
Linux and Mac users can get along pretty well. Apple (setting aside Quicktime) doesn't push proprietary formats, a la Microsoft. Apple doesn't play dirty compatibility games, a la Microsoft ("Gee, did we break Netscape Server with that change? Do we prioritize IE requiests ahead of Navigator requests? Ooops, looks like we introduced a bug!"). An Apple/Linux argument is much like an argument between vi and emacs. It can get very impassioned, as each person defends their own favorite logic. However, in the end, the two interoperate well -- they'e still both churning out text. Bob in the next cubicle can use one, and me the other, and everyone is happy. I don't get my nice GNU tools on a vanilla Mac, but I get a reasonable set of POSIX utils. I can write and run my scripts and work without too much pain. I don't have the Godawful Windows virtual terminal and horrendous shell. I get X11 support. Yeah, some Linux software doesn't work well or at all under OS X (especially for things that have half-done native ports from X11), and software using the Mac's GUI as a front-end doesn't work really well on Linux. However, think of the following:
* Libraries can be designed to be cross-platform. Most Windows uers that I know of seem to use AIM or ICQ, or maybe Trillian, which I believe is a closed-source codebase. The friend that I have that uses OS X uses Adium, which uses libgaim. If I find a bug in libgaim on my Linux box and fix it, he benefits, and visa versa. There are a startling number of Mac OS X people working on POSIX sourceforge projects, much like the Linux world, and very unlike the Windows world.
* Dunno if Mac OS X does perl out of box, but if not, I'm sure that it's installable via fink or something. I don't have to futz with Visual Basic crap coming from some annoying Windows "programmer". Similarly, nice traditional UNIX C daemons work nicely on OS X *or* Linux.
* Objective C. Linux has a nice Objective C compiler available in the GNU Compiler Collection that ships with most distros. The only guy I know that uses Objective C isn't really impressed with it, but still, if you like using the language of choice on the Mac, you can code on Linux comfortably.
* X11 support. I can run X11 apps anywhere, and over the network. It's a whole different world from Windows.
So, while people may happily bash someone else's OS between Mac OS X and Linux, ultimately they can live together pretty comfortably. I mean, I get really annoyed when using Solaris, which I see as missing features, being heavyweight, and being rather expensive. However, I can do useful work on Solaris without constantly getting ticked off at having an environment about a tenth as capable as my Linux box at home -- which is exactly what happens when I use a Windows machine.
Quite frankly, for an entire generation of users, if it can't sync with an iPod, it's not for general use.
Wow. Someone that purchases their computer based on their Walkman replacement? Hell, I'd imagine that most people do the opposite -- buy an MP3 player that has support for their computer. Well, in any event, for folks that like iPods, it's quite fortunate that Linux works with the iPod. Hell, my Linux box works with iPods and my Mac next to it, an old system running OS 8, doesn't.
I know of several people using first generation PPC machines, simply because they do everything needed.
6100/7100/8100 boxes? Christ, I had a 6100 with a 60Mhz processor, and I just got a 180Mhz PowerPC machine (three *times* the speed) with 17" monitor, keyboard, mouse, and the works for $30 a couple of months ago. I don't think I'd bother with any first-gen Power Macs any more -- you simply cannot reasonably browse the web on one, which is a pretty major limitation for most folks.
I admit that I still buy hardware with one eye open for Linux -- if I'm getting something like a webcam or scanner, I'm going to do a quick google to make sure that someone's used it before on Linux.
However, the situation seems to be pretty solid. I have an X10 controller (not the sort of thing that most folk have around) that doesn't ship with drivers with a stock Red Hat kernel, but everything else works out of box.
On the other hand, I *have* had multiple instances of hardware that *did* work on Linux just fine failing on new releases of Windows. This usually happens when a vendor goes out of business. They have nobody left producing new drivers (hence new versisions of Windows get new drivers). Linux systems, with open-source drivers, get patched and maintained as time goes on. I've had a friend with an Aureal sound card that worked in Linux but not Windows and an old oddball soundcard at work that worked in Linux but hadn't had any Windows drivers for the NT/2k/XP line.
Game support is still not there, though -- most genres of games don't do too well with open-source projects, and so there isn't a lot out there, which means that there aren't a lot of Linux gamers, which means that few commercial companies do Linux releases.
I'd like to point out that every single major alternative to Windows is now UNIX-y, at the least. Dunno about the details of BeOS, though it's supposed to be at least POSIX compliant. Mac OS X is BSDish. BSD is UNIX. Linux is UNIXy. Solaris is UNIX. The number of people using UNIXy systems is shooting up each day.:-) Mmmm....powerful tools everywhere.
DVD player: mplayer, xine, videolan, just name a few. CD writer: xcdroast, gtoaster, k3b, not counting commandline apps. They either come with your distro, or are really easy to get. Apt-get, rpmfind.net, emerge, you name it.
Third-party people package these, but Red Hat has the biggest market share out there, and their latest release, Fedora Core 1, does not provide mplayer, xine, or videolan in the default repository. True, other people package these for FC1, but they aren't in your yum.conf by default. MPlayer in particular was, in the past, a real bear to build ahead of time and distribute as a binary, because there were a ton of arch-specific build-time optimizations. There are also legal concerns associated with a number of formats -- what *exactly* is the legal status of DiVX, given that it might use IP from the MPEG folks? How about MP3 (another high-profile omission from RH9's xmms due to legal reasons, when the Fraunhoffer people were blustering about lawsuits, though it's back in FC1)? What about Sorenson support, or the Windows binaries needed to work with a number of proprietary codecs? In the past, a lot of gray area has been ignored by little projects poking around on the Internet, but you can't have a distro maintainer ignore things -- they have pocketbooks that one can sue for.
Also, CD burning used to be a pain in the ass (not sure what the status is in FC1) in Red Hat, because RH used the ide-cd driver by default...and for a long time, you couldn't burn CDs without using ide-scsi emulation.
Note that other vendors have had flaws as well -- I've been using Red Hat's releases for *years*, and can dredge a fair number of irritations folks have had up pretty easily.
This one is a no no, for two reasons: (a) If a windows software is useful, we better have a native version, for the sake of reliability/performance/freedom. Write a clone or have the vender port it, just don't count on running an emulator as a long term solution. (b) If you have to use Wine to get something done, what would a new/non-export/prospect user--which seems to be your major concern--think? "Look, linux must suck, because it doesn't even have that!"
Right. WINE performance is actually not bad (assumine one has it set up properly). Freedom is an issue. WINE reliability is a decided issue -- you can't just have half of all Windows software work, and the other part fail in random ways or during certain tasks. That's unacceptable for an end user. It's nice for a techie who can't get away with using Linux without a particular Windows app, and has been valuable for the past few years, but users can start expecting native releases. WINE changes not infrequently break previously working software -- it's a tough job to reverse engineer and implement a huge chunk of buggy software like the collection of Windows libraries.
Not sure what you refer to. Maybe you should have named some Gnome/KDE default setting that's wacky? On a side note, "usability" is overrated. Gnome's default wm, metacity, is a prime "usability" example, made by "usability" export Havoc Pennington after consulting "usability" surveys, and it ends up being one of the most unusable piece of software in the Gnome packages. If this happens to be what you meant "improvements and surveys will help", future looks quite dark to me.
I agree. Having an easy-to-use environment out-of-box has not been a problem for years on Red Hat. I can see folks complaining when you got an ancient AfterStep desktop that required a significant amount of config file editing and WM restarting to make even basic changes. This is long behind us.
This totally depends on what you do with you desktop. Try come up with a GUI that can do things such as putting ".virus" suffix on all.eml files scattered all over you directory tree. CLI does it easy.
People with no administrative responsibilities do not, IMHO, have to use the CLI. People with any administrative tasks d
1. Good DVD player & CD-RW that just work, without mesing around. If this software is not part of the distro, simple instructions on how to get/install it (one click?).
Fair enough. MPlayer and xine are both good (though MPlayer doesn't have a WMP-style GUI that's usable in fullscreen mode, and may put some folks off), but I know that currently neither is part of out-of-box Fedora.
2. Friends who are familiar with the OS/Distro, for the network effects and piece of mind in case something goes drastically wrong. This is where having a "critical mass" (fuzzy value) comes in - this is already happening, but the more, the better.
Mmmmf. I dunno about this. I agree that it could be higher, but I find that the same techie people that get asked Windows questions tend to also use Linux.
3. Better Wine, but that will come with age.:) [CodeWeavers is doing a really good job, with full disclosure of the limitations, which leads to a sense of psychological well being, rather than the feeling "they are trying to take advantage of me."]
I'd modify that to "better Linux binary packaging". It's a bitch to package a single binary that runs on many Linux distros, and there's no approach that allows a user a consistent install interface. Native apps will come -- I don't think WINE can be a long-term solution, since "works most of the time" isn't really good enough for most users, but they *need* to be as easy to package and with the backwards compatibility that Windows has.
4. Better default settings for Desktop/Window managers that make sense to a majority (and keep the ability to tweak). The "usability" improvements and surveys will help here, a lot. More needs to happen in that field.
I can't agree. Everyone that has used a current Linux distro that I know of was pretty comfortable. There's a bit of exploration time, while folks find the equivalent of "Control Panels" and the like, but out-of-box, things tend to be pretty easy to use. I personally think that the defaults are toned down enough that most power users should blow away the default KDE/GNOME WMs, as neither is nearly as powerful as some of the other options available.
5. Use easier "language" - eventually (in 1-2 years) e.g., non-cryptic commands, or a *standardized* set of aliases that work on all distros. [And continue to evolve the GUI so the user doesn't HAVE TO use the CLI.]
I cannot agree. A Joe User office type does not have to use the CLI. Someone with any sort of administrative responisibilities does. As for aliases, no. Windows did fine with "del" and "type", and UNIX clones are fine with "rm" and "cat". Forcing everyone to learn a new set of aliases is ridiculous.
6. Better Grub/Lilo/equivalent that is less intimidating for new users that want multi-boot. Preferably with a easy to use GUI that detects all HDDs & partitions and tells you what's on them (with as much relevant information as possible).
Done in the installers for major distros. I don't think current users need to poke at grub.conf or lilo.conf at all to multi-boot with Windows.
7. Some packaging system with less dependency problems. [Yes, there are a few that show very good promise, with only occasional issues surfacing.]
I cannot agree. Dependency problems are *not* the issue -- all current major distros have auto-download and install systems that handle dependencies. The problem is that it's still difficult to build and package a single binary that will run on multiple platforms, and until that is the case, people will *continue* to try to run, say, Red Hat packages on Mandrake and become frusterated when problems pop up. That is not an issue for a packaging system alone.
9. Few, well chosen default applications on the distro (not "give them four of everything"). [Lot of progress has already happened in this area in a few distros.]
I know few people using Linux at home. Linux is being deployed as a business desktop, a cubicle box, which was the area traditionally ruled by Microsoft. Apple's generally stayed with niches.
The fact is, you can *combine* the Apple and Linux desktop market share to calculate (desktop UNIX users) and watch happily as it rises. Mmmm....
You bring up an interesting point. What if the Windows source was analyzed for stolen GPL code? Wouldn't it make that part of the Windows OS GPL?
No. It *would* mean that Microsoft had infringed on the original author's copyrights -- they never had rights to use the code in the first place.
You can never "accidently GPL" something you wrote by simply combining your (GPL-incompatible) code with GPL code. However, if you do so, and your code is not GPL-compatible, you *are* liable for copyright infringement.
I don't think the FSF would go after Microsft for infringement unless the infringement was pretty severe or obviously deliberate. The FSF has been pretty reasonable in the past with GPL infringements -- they want to discourage folks from stealing code, but neither are they ambulance-chasers, trying to get multi-bazallion dollar settlements.
No. Entrapment requires that someone be incited to commit a crime. Simply recording what someone does is not illegal, since the loggers are not directly encouraging someone to commit a crime.
If it is USB and HID-compliant, it probably already works (note that using many-button USB mice is a pain to set up -- here's my XF86Config section for a five-button (plus up and down on a scrollwheel, which is not properly detected by Fedora) config:
Feel free to release Verilog designs under a GPL-like license and start a revolution. Note that it took Stallman and Co. years of work to get much notice.
No. If the Wine folks look at the actual Windows source code, they aren't reverse engineering any more, they're copying, which is illegal.
IANAL. You are wrong. Non-clean-room reverse engineering is not only legal but is done at many, many companies. There is *absolutely no constraint* to use a clean room in reverse engineering.
The first clean room reverse engineering that I'm aware of is Phoenix of IBM's BIOS. They had *no* legal requirement to clean-room reverse engineer the BIOS. If they wanted to, they could hire IBM BIOS engineers for the job. However, by doing a clean room implementation, they ensured that they had an counterargument to *any* potential IBM claims of infringement. Had they not have used a cleanroom tactic, they might have had to actually have folks look at the code and at what people were doing with the code if charged with infringement. While this can be useful -- it's an immediate shutdown to any argument IBM might raise about infringement in court, and the judge doesn't even need to see the code -- it is definitely not necessary. I can look at GPL code and use the same approach said code does as long as I am not copying code verbatim (note that changing variables or something is not sufficient -- the work must be done by you, not be a mangled version of the original).
That being said, WINE has long had a policy of *not* accepting access to Windows source code. They've had people with access to it volunteer to give them stuff in the past, and they want to do a pseudo-cleanroom approach, since it makes matters simple from a legal standpoint. WINE will probably continue to ignore the source (and the WINE maintainers now have to worry about people submitting WINE patches containing Windows source...they may require indemnification or God knows what).
From a security standpoint, this is an utter disaster to Microsoft. They haven't had the benefit of many eyes all these years, and now they have a fucking lot of malicious eyes, and ten years of holes to remove in a week or so before the nastier exploits come out. None of those eyes have any incentive to submit patches to Microsoft. There will be attacks on relatively hardened systems, too.
This is going to suck for friends and family that I have using Windows.
You have a 1GB (I assume not giga*bit*) limit on traffic per month?
Dump your ISP.
Even someone saturating a 56K modem can download about 12GB a month. You're getting shafted, unless you're paying less than a tenth of what a dialup account runs.
ISP bandwidth costs, based on colo prices I've seen at servercove.com, are no more than $99 for 700GB of data transferred. Given that you're probably paying at least $30 a month for broadband...
I can't debate the Kerberos packages, as I haven't ever used . Back when I was setting up Kerberos on RH, *nobody* had *any* Kerberos support (especially for Kerb 4) and you had to roll a lot of your own stuff. It sucked. A lot. I dunno what RH's done wrong with Kerb, though. I'm not sure what packages you're upset with static linking in -- I don't see a whole lot of statically linked packages.
On the other hand, if you don't like RH 9, it's certainly reasonable to assume that you may well not like Fedora Core 1. My beef is with people that used and liked the Red Hat series, but then got the idea that the Fedora Core series was anything different. It's *exactly* the same distro, plus some extra packages and with a name change and is produced by exactly the same folks at Red Hat.
I think that if I had to summarize the reason Marathon's story made such a huge impression on everyone (and spawned the massive Marathon Story) site and many thousands of posts of sophisticated discussion in two words, I'd use "subtle" and "deep".
Marathon made a number of very subtle, minor allusions, clues to figuring out the full story. It did not come out and simply say "Durandal is a bad computer who is insane. You need to work with him." or something like that. It let you discover the plot as you went through tidbits dropped. You couldn't just read the story by going through the terminals and appending one to another -- each is a non-chronologically sequential piece of information that generally fits into the story somehow. It's quite Myth-like, though with a far more complex story.
Marathon's plot is very deep for such a subtle one -- there are a *phenomenal* number of references to literature, history, mythology, contemporary weapons technology, psychology, etc, that you need to look up to understand fine points of what the authors were getting at. You might find yourself reading a news article on the sociology of a Martian political group that gives some insight into the background of what happend and the reason things occurred. Because it took such expertise to figure the story out, it spawned a vast number of (sometimes occasional) players who wrote interesting analyses from their own area of expertise to help the community figure out what Bungie had written.
Even better, you didn't *have* to pay attention to the story if you didn't want to. You weren't forced to rely on the story much to figure out what to do in the game. You could play Marathon as a straight action game if you wanted to, and weren't interested in the story, but if you wanted to get into the story, it was appealing and there. This made the game appealing to a broad audience.
If I could choose two more elements, probably less important, that made Marathon good, it would be darkness and plausibility.
The Marathon comments were frequently very dark (a trend that progressed as the series continued). They referred to deaths and killings quite seriously. They referenced massacres and insanity, and not in a offhandish way in the least. They also did not generally say "someone is insane" -- they let you figure it out for yourself, by reading their thoughts or what happened. One of the darkest is the infamous Gheritt White terminal. This is one of the darkest and most disturbing texts that I think I've ever read in a video game -- much more intimidating than the short and violence-glorifying snippits in Postal. That single terminal alone spawned *vast* amounts of discussion and analysis. When elements in Marathon II and Infinity (like the pocketknife/broadsword terminal) start their own story threads that start out reasonable and get darker, you can really feel a kind of shocked surprise. If you're playing by yourself, late at night and in the dark, (and have just survived creeping through dark hallways with silent things drifting down them and around corners and out of the darkness) your words are probably much like mine -- "Oh, *man*". The only games that I think have competed with Marathon in terms of slowly, horrifyingly uncovering what happened are adventure games -- like Myth -- and I've yet to see an adventure game with the subtlety of Marathon.
Marathon is also plausible. There were, to be fair, errors. However, Marathon's story underwent the most extensive analysis I've ever seen a story undergo. I doubt that books undergo such work, especially given the size of the crowd looking for errors. Like Snow Crash, much of the computer technology in the series is at least acceptably plausible. Real terms are used, references to current technologies are used. It means that programmers don't have to constantly wince when playing the game, which is truly wonderful and unusual for a st
I dunno. I'm not a LoTR expert, but I'd say that Sauron isn't even much of a character -- he's kind of an abstract, a general badness. He never even personally *enters* the story. The only thing that really comes in is his gaze or a few remote words that act in kind of a vague hand-of-God mode -- you don't understand God's character from him lighting a couple of piles of soaked logs on fire in the Bible, you know? You don't expect the Devil to have character development, even if he exerts energies in a book, and Sauron plays a similar role. IMHO, of course.
Saruman, I think, was not a flat evil character. He was clearly good at one point, and got too ambitious. Even then, he wasn't exactly evil -- he even tried to peacefully recruit Gandalf at one point, with the goal of going after Sauron, not becoming ruler of the world. He was kind of like Boromir -- he chose the wrong method to achive his goals, and ended up sinking deeper and deeper.
Max is interesting, but remember that he isn't actually all that complex.
The idea of much of the content in Max Payne is to build up a stereotypical character from film noir/comic books. It's not unenjoyable, just that it shouldn't be confused with being all that sophisticated.
The reason we see Max Payne as "fairly good" is that most video games are "fairly bad" when it comes to characters.
Remember what TV series frequently do with multiple episode story arcs? They have a quick introduction to where the story left off when they resume. It'd be pretty easy to include an option to "brief me on the story". When you combine that with a Neverwinter Nights-style journal for more complex games and the ability to "look" or "get info" on something or someone and see a description and things that you've learned about it, you can pretty much stop and go whenever you want.
Remember that lots of people read, say, a chapter of a book a night and still enjoy it, and a book can't do the kind of "quick recaps" that TV episodes do and video games *could* do.
The metric that video game plots and characters are judged by is wildly generous when compared to, say, movie equivalents. I hear people all the time saying that "Final Fantasy games have good plots" or "Final Fantasy games have good characters". Well, dammit, Final Fantasy really, really sucks in *both* areas compared to a good movie.
Some of that is because lots of games are intended to be played by children, and game authors (Nintendo produces a lot of these) find that it's easier to present folks with a very clear good guy and another very clear bad guy, maybe making the bad guy a bit goofy or suffering a number of character flaws and weaknesses.
Game budgets are big enough that Japanese-to-English translation and other technical obstacles of video games should not be a barrier anymore. We can expect *good* writing (good by *book* standards -- they have to write a lot less text and get a lot more money), engaging characters (same reason), and good plots. *Not* "good plot for a video game", which is what "good plot" has conventionally meant when discussing video games.
I remember when people went absolutely nuts when a video game actually included -- here it comes -- a *plot twist*! Awkward pauses in speech and unreadable expressions were also par for the course. Developers *can* do better than this, and *should*.
Slashdot Mirror
I tend not to go to the sort of sites that do this, but I'd look for "promo" sites, the kind of thing that push a product and tend to have lots of annoying animation and music. The sort of site that's generally a pain in the ass to use.
At one point (not still the case) the official Star Trek site complained about anything other than IE. Sites for new movies are another good bet.
Good luck on Linux overtaking OS X's momentum.
Apple has built themselves a very profitable niche, and is happy with it. Yes, they could get more market share (but much less money, at least in the short run) by porting Mac OS X to x86. They've chosen not to. They want to maintain a market with users who are comfortable paying high prices for a polished black box system. Nothing wrong with that, but Apple is not aiming at the masses, which buy computers based on price, where Apple simply is not competitive (and, again, has chosen not to be).
Aside from the folks that use Linux on PowerPC, Apple's Mac OS doesn't really even compete with Linux all that much. People using Linux on the desktop are generally on x86 -- a new Linux user means a vanished Windows users.
Linux and Mac users can get along pretty well. Apple (setting aside Quicktime) doesn't push proprietary formats, a la Microsoft. Apple doesn't play dirty compatibility games, a la Microsoft ("Gee, did we break Netscape Server with that change? Do we prioritize IE requiests ahead of Navigator requests? Ooops, looks like we introduced a bug!"). An Apple/Linux argument is much like an argument between vi and emacs. It can get very impassioned, as each person defends their own favorite logic. However, in the end, the two interoperate well -- they'e still both churning out text. Bob in the next cubicle can use one, and me the other, and everyone is happy. I don't get my nice GNU tools on a vanilla Mac, but I get a reasonable set of POSIX utils. I can write and run my scripts and work without too much pain. I don't have the Godawful Windows virtual terminal and horrendous shell. I get X11 support. Yeah, some Linux software doesn't work well or at all under OS X (especially for things that have half-done native ports from X11), and software using the Mac's GUI as a front-end doesn't work really well on Linux. However, think of the following:
* Libraries can be designed to be cross-platform. Most Windows uers that I know of seem to use AIM or ICQ, or maybe Trillian, which I believe is a closed-source codebase. The friend that I have that uses OS X uses Adium, which uses libgaim. If I find a bug in libgaim on my Linux box and fix it, he benefits, and visa versa. There are a startling number of Mac OS X people working on POSIX sourceforge projects, much like the Linux world, and very unlike the Windows world.
* Dunno if Mac OS X does perl out of box, but if not, I'm sure that it's installable via fink or something. I don't have to futz with Visual Basic crap coming from some annoying Windows "programmer". Similarly, nice traditional UNIX C daemons work nicely on OS X *or* Linux.
* Objective C. Linux has a nice Objective C compiler available in the GNU Compiler Collection that ships with most distros. The only guy I know that uses Objective C isn't really impressed with it, but still, if you like using the language of choice on the Mac, you can code on Linux comfortably.
* X11 support. I can run X11 apps anywhere, and over the network. It's a whole different world from Windows.
So, while people may happily bash someone else's OS between Mac OS X and Linux, ultimately they can live together pretty comfortably. I mean, I get really annoyed when using Solaris, which I see as missing features, being heavyweight, and being rather expensive. However, I can do useful work on Solaris without constantly getting ticked off at having an environment about a tenth as capable as my Linux box at home -- which is exactly what happens when I use a Windows machine.
Ugh. You may be right, and I love my Linux machine, but hearing it being claimed that Windows support is a killer app makes my stomach roil.
Quite frankly, for an entire generation of users, if it can't sync with an iPod, it's not for general use.
Wow. Someone that purchases their computer based on their Walkman replacement? Hell, I'd imagine that most people do the opposite -- buy an MP3 player that has support for their computer. Well, in any event, for folks that like iPods, it's quite fortunate that Linux works with the iPod. Hell, my Linux box works with iPods and my Mac next to it, an old system running OS 8, doesn't.
I know of several people using first generation PPC machines, simply because they do everything needed.
6100/7100/8100 boxes? Christ, I had a 6100 with a 60Mhz processor, and I just got a 180Mhz PowerPC machine (three *times* the speed) with 17" monitor, keyboard, mouse, and the works for $30 a couple of months ago. I don't think I'd bother with any first-gen Power Macs any more -- you simply cannot reasonably browse the web on one, which is a pretty major limitation for most folks.
I admit that I still buy hardware with one eye open for Linux -- if I'm getting something like a webcam or scanner, I'm going to do a quick google to make sure that someone's used it before on Linux.
However, the situation seems to be pretty solid. I have an X10 controller (not the sort of thing that most folk have around) that doesn't ship with drivers with a stock Red Hat kernel, but everything else works out of box.
On the other hand, I *have* had multiple instances of hardware that *did* work on Linux just fine failing on new releases of Windows. This usually happens when a vendor goes out of business. They have nobody left producing new drivers (hence new versisions of Windows get new drivers). Linux systems, with open-source drivers, get patched and maintained as time goes on. I've had a friend with an Aureal sound card that worked in Linux but not Windows and an old oddball soundcard at work that worked in Linux but hadn't had any Windows drivers for the NT/2k/XP line.
Game support is still not there, though -- most genres of games don't do too well with open-source projects, and so there isn't a lot out there, which means that there aren't a lot of Linux gamers, which means that few commercial companies do Linux releases.
I'd like to point out that every single major alternative to Windows is now UNIX-y, at the least. Dunno about the details of BeOS, though it's supposed to be at least POSIX compliant. Mac OS X is BSDish. BSD is UNIX. Linux is UNIXy. Solaris is UNIX. The number of people using UNIXy systems is shooting up each day. :-) Mmmm....powerful tools everywhere.
DVD player: mplayer, xine, videolan, just name a few.
.eml files scattered all over you directory tree. CLI does it easy.
CD writer: xcdroast, gtoaster, k3b, not counting commandline apps.
They either come with your distro, or are really easy to get. Apt-get, rpmfind.net, emerge, you name it.
Third-party people package these, but Red Hat has the biggest market share out there, and their latest release, Fedora Core 1, does not provide mplayer, xine, or videolan in the default repository. True, other people package these for FC1, but they aren't in your yum.conf by default. MPlayer in particular was, in the past, a real bear to build ahead of time and distribute as a binary, because there were a ton of arch-specific build-time optimizations. There are also legal concerns associated with a number of formats -- what *exactly* is the legal status of DiVX, given that it might use IP from the MPEG folks? How about MP3 (another high-profile omission from RH9's xmms due to legal reasons, when the Fraunhoffer people were blustering about lawsuits, though it's back in FC1)? What about Sorenson support, or the Windows binaries needed to work with a number of proprietary codecs? In the past, a lot of gray area has been ignored by little projects poking around on the Internet, but you can't have a distro maintainer ignore things -- they have pocketbooks that one can sue for.
Also, CD burning used to be a pain in the ass (not sure what the status is in FC1) in Red Hat, because RH used the ide-cd driver by default...and for a long time, you couldn't burn CDs without using ide-scsi emulation.
Note that other vendors have had flaws as well -- I've been using Red Hat's releases for *years*, and can dredge a fair number of irritations folks have had up pretty easily.
This one is a no no, for two reasons: (a) If a windows software is useful, we better have a native version, for the sake of reliability/performance/freedom. Write a clone or have the vender port it, just don't count on running an emulator as a long term solution. (b) If you have to use Wine to get something done, what would a new/non-export/prospect user--which seems to be your major concern--think? "Look, linux must suck, because it doesn't even have that!"
Right. WINE performance is actually not bad (assumine one has it set up properly). Freedom is an issue. WINE reliability is a decided issue -- you can't just have half of all Windows software work, and the other part fail in random ways or during certain tasks. That's unacceptable for an end user. It's nice for a techie who can't get away with using Linux without a particular Windows app, and has been valuable for the past few years, but users can start expecting native releases. WINE changes not infrequently break previously working software -- it's a tough job to reverse engineer and implement a huge chunk of buggy software like the collection of Windows libraries.
Not sure what you refer to. Maybe you should have named some Gnome/KDE default setting that's wacky? On a side note, "usability" is overrated. Gnome's default wm, metacity, is a prime "usability" example, made by "usability" export Havoc Pennington after consulting "usability" surveys, and it ends up being one of the most unusable piece of software in the Gnome packages. If this happens to be what you meant "improvements and surveys will help", future looks quite dark to me.
I agree. Having an easy-to-use environment out-of-box has not been a problem for years on Red Hat. I can see folks complaining when you got an ancient AfterStep desktop that required a significant amount of config file editing and WM restarting to make even basic changes. This is long behind us.
This totally depends on what you do with you desktop. Try come up with a GUI that can do things such as putting ".virus" suffix on all
People with no administrative responsibilities do not, IMHO, have to use the CLI. People with any administrative tasks d
1. Good DVD player & CD-RW that just work, without mesing around. If this software is not part of the distro, simple instructions on how to get/install it (one click?).
Fair enough. MPlayer and xine are both good (though MPlayer doesn't have a WMP-style GUI that's usable in fullscreen mode, and may put some folks off), but I know that currently neither is part of out-of-box Fedora.
2. Friends who are familiar with the OS/Distro, for the network effects and piece of mind in case something goes drastically wrong. This is where having a "critical mass" (fuzzy value) comes in - this is already happening, but the more, the better.
Mmmmf. I dunno about this. I agree that it could be higher, but I find that the same techie people that get asked Windows questions tend to also use Linux.
3. Better Wine, but that will come with age.
I'd modify that to "better Linux binary packaging". It's a bitch to package a single binary that runs on many Linux distros, and there's no approach that allows a user a consistent install interface. Native apps will come -- I don't think WINE can be a long-term solution, since "works most of the time" isn't really good enough for most users, but they *need* to be as easy to package and with the backwards compatibility that Windows has.
4. Better default settings for Desktop/Window managers that make sense to a majority (and keep the ability to tweak). The "usability" improvements and surveys will help here, a lot. More needs to happen in that field.
I can't agree. Everyone that has used a current Linux distro that I know of was pretty comfortable. There's a bit of exploration time, while folks find the equivalent of "Control Panels" and the like, but out-of-box, things tend to be pretty easy to use. I personally think that the defaults are toned down enough that most power users should blow away the default KDE/GNOME WMs, as neither is nearly as powerful as some of the other options available.
5. Use easier "language" - eventually (in 1-2 years) e.g., non-cryptic commands, or a *standardized* set of aliases that work on all distros. [And continue to evolve the GUI so the user doesn't HAVE TO use the CLI.]
I cannot agree. A Joe User office type does not have to use the CLI. Someone with any sort of administrative responisibilities does. As for aliases, no. Windows did fine with "del" and "type", and UNIX clones are fine with "rm" and "cat". Forcing everyone to learn a new set of aliases is ridiculous.
6. Better Grub/Lilo/equivalent that is less intimidating for new users that want multi-boot. Preferably with a easy to use GUI that detects all HDDs & partitions and tells you what's on them (with as much relevant information as possible).
Done in the installers for major distros. I don't think current users need to poke at grub.conf or lilo.conf at all to multi-boot with Windows.
7. Some packaging system with less dependency problems. [Yes, there are a few that show very good promise, with only occasional issues surfacing.]
I cannot agree. Dependency problems are *not* the issue -- all current major distros have auto-download and install systems that handle dependencies. The problem is that it's still difficult to build and package a single binary that will run on multiple platforms, and until that is the case, people will *continue* to try to run, say, Red Hat packages on Mandrake and become frusterated when problems pop up. That is not an issue for a packaging system alone.
9. Few, well chosen default applications on the distro (not "give them four of everything"). [Lot of progress has already happened in this area in a few distros.]
No. That would be incredibly agg
I know few people using Linux at home. Linux is being deployed as a business desktop, a cubicle box, which was the area traditionally ruled by Microsoft. Apple's generally stayed with niches.
The fact is, you can *combine* the Apple and Linux desktop market share to calculate (desktop UNIX users) and watch happily as it rises. Mmmm....
You bring up an interesting point. What if the Windows source was analyzed for stolen GPL code? Wouldn't it make that part of the Windows OS GPL?
No. It *would* mean that Microsoft had infringed on the original author's copyrights -- they never had rights to use the code in the first place.
You can never "accidently GPL" something you wrote by simply combining your (GPL-incompatible) code with GPL code. However, if you do so, and your code is not GPL-compatible, you *are* liable for copyright infringement.
I don't think the FSF would go after Microsft for infringement unless the infringement was pretty severe or obviously deliberate. The FSF has been pretty reasonable in the past with GPL infringements -- they want to discourage folks from stealing code, but neither are they ambulance-chasers, trying to get multi-bazallion dollar settlements.
No. Entrapment requires that someone be incited to commit a crime. Simply recording what someone does is not illegal, since the loggers are not directly encouraging someone to commit a crime.
If it is USB and HID-compliant, it probably already works (note that using many-button USB mice is a pain to set up -- here's my XF86Config section for a five-button (plus up and down on a scrollwheel, which is not properly detected by Fedora) config:
Section "InputDevice"
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "ExplorerPS/2"
Option "Device" "/dev/mouse"
Option "Buttons" "7"
Option "ZAxisMapping" "6 7"
Option "SendCoreEvents" "true"
EndSection
I believe that this is a pain in the ass with PS/2 mice, as the protocols can differ.
Feel free to release Verilog designs under a GPL-like license and start a revolution. Note that it took Stallman and Co. years of work to get much notice.
Photoshop did not exist in 1969.
Do you want NASA to give away its latest and greatest rocket guidance software to anyone that asks for it?
Yes.
Argh. Bungie on the brain. Substitute "Myst" for "Myth" in the article.
No. If the Wine folks look at the actual Windows source code, they aren't reverse engineering any more, they're copying, which is illegal.
IANAL. You are wrong. Non-clean-room reverse engineering is not only legal but is done at many, many companies. There is *absolutely no constraint* to use a clean room in reverse engineering.
The first clean room reverse engineering that I'm aware of is Phoenix of IBM's BIOS. They had *no* legal requirement to clean-room reverse engineer the BIOS. If they wanted to, they could hire IBM BIOS engineers for the job. However, by doing a clean room implementation, they ensured that they had an counterargument to *any* potential IBM claims of infringement. Had they not have used a cleanroom tactic, they might have had to actually have folks look at the code and at what people were doing with the code if charged with infringement. While this can be useful -- it's an immediate shutdown to any argument IBM might raise about infringement in court, and the judge doesn't even need to see the code -- it is definitely not necessary. I can look at GPL code and use the same approach said code does as long as I am not copying code verbatim (note that changing variables or something is not sufficient -- the work must be done by you, not be a mangled version of the original).
That being said, WINE has long had a policy of *not* accepting access to Windows source code. They've had people with access to it volunteer to give them stuff in the past, and they want to do a pseudo-cleanroom approach, since it makes matters simple from a legal standpoint. WINE will probably continue to ignore the source (and the WINE maintainers now have to worry about people submitting WINE patches containing Windows source...they may require indemnification or God knows what).
From a security standpoint, this is an utter disaster to Microsoft. They haven't had the benefit of many eyes all these years, and now they have a fucking lot of malicious eyes, and ten years of holes to remove in a week or so before the nastier exploits come out. None of those eyes have any incentive to submit patches to Microsoft. There will be attacks on relatively hardened systems, too.
This is going to suck for friends and family that I have using Windows.
You have a 1GB (I assume not giga*bit*) limit on traffic per month?
Dump your ISP.
Even someone saturating a 56K modem can download about 12GB a month. You're getting shafted, unless you're paying less than a tenth of what a dialup account runs.
ISP bandwidth costs, based on colo prices I've seen at servercove.com, are no more than $99 for 700GB of data transferred. Given that you're probably paying at least $30 a month for broadband...
I can't debate the Kerberos packages, as I haven't ever used
. Back when I was setting up Kerberos on RH, *nobody* had *any* Kerberos support (especially for Kerb 4) and you had to roll a lot of your own stuff. It sucked. A lot. I dunno what RH's done wrong with Kerb, though. I'm not sure what packages you're upset with static linking in -- I don't see a whole lot of statically linked packages.
On the other hand, if you don't like RH 9, it's certainly reasonable to assume that you may well not like Fedora Core 1. My beef is with people that used and liked the Red Hat series, but then got the idea that the Fedora Core series was anything different. It's *exactly* the same distro, plus some extra packages and with a name change and is produced by exactly the same folks at Red Hat.
I think that if I had to summarize the reason Marathon's story made such a huge impression on everyone (and spawned the massive Marathon Story) site and many thousands of posts of sophisticated discussion in two words, I'd use "subtle" and "deep".
Marathon made a number of very subtle, minor allusions, clues to figuring out the full story. It did not come out and simply say "Durandal is a bad computer who is insane. You need to work with him." or something like that. It let you discover the plot as you went through tidbits dropped. You couldn't just read the story by going through the terminals and appending one to another -- each is a non-chronologically sequential piece of information that generally fits into the story somehow. It's quite Myth-like, though with a far more complex story.
Marathon's plot is very deep for such a subtle one -- there are a *phenomenal* number of references to literature, history, mythology, contemporary weapons technology, psychology, etc, that you need to look up to understand fine points of what the authors were getting at. You might find yourself reading a news article on the sociology of a Martian political group that gives some insight into the background of what happend and the reason things occurred. Because it took such expertise to figure the story out, it spawned a vast number of (sometimes occasional) players who wrote interesting analyses from their own area of expertise to help the community figure out what Bungie had written.
Even better, you didn't *have* to pay attention to the story if you didn't want to. You weren't forced to rely on the story much to figure out what to do in the game. You could play Marathon as a straight action game if you wanted to, and weren't interested in the story, but if you wanted to get into the story, it was appealing and there. This made the game appealing to a broad audience.
If I could choose two more elements, probably less important, that made Marathon good, it would be darkness and plausibility.
The Marathon comments were frequently very dark (a trend that progressed as the series continued). They referred to deaths and killings quite seriously. They referenced massacres and insanity, and not in a offhandish way in the least. They also did not generally say "someone is insane" -- they let you figure it out for yourself, by reading their thoughts or what happened. One of the darkest is the infamous Gheritt White terminal. This is one of the darkest and most disturbing texts that I think I've ever read in a video game -- much more intimidating than the short and violence-glorifying snippits in Postal. That single terminal alone spawned *vast* amounts of discussion and analysis. When elements in Marathon II and Infinity (like the pocketknife/broadsword terminal) start their own story threads that start out reasonable and get darker, you can really feel a kind of shocked surprise. If you're playing by yourself, late at night and in the dark, (and have just survived creeping through dark hallways with silent things drifting down them and around corners and out of the darkness) your words are probably much like mine -- "Oh, *man*". The only games that I think have competed with Marathon in terms of slowly, horrifyingly uncovering what happened are adventure games -- like Myth -- and I've yet to see an adventure game with the subtlety of Marathon.
Marathon is also plausible. There were, to be fair, errors. However, Marathon's story underwent the most extensive analysis I've ever seen a story undergo. I doubt that books undergo such work, especially given the size of the crowd looking for errors. Like Snow Crash, much of the computer technology in the series is at least acceptably plausible. Real terms are used, references to current technologies are used. It means that programmers don't have to constantly wince when playing the game, which is truly wonderful and unusual for a st
I dunno. I'm not a LoTR expert, but I'd say that Sauron isn't even much of a character -- he's kind of an abstract, a general badness. He never even personally *enters* the story. The only thing that really comes in is his gaze or a few remote words that act in kind of a vague hand-of-God mode -- you don't understand God's character from him lighting a couple of piles of soaked logs on fire in the Bible, you know? You don't expect the Devil to have character development, even if he exerts energies in a book, and Sauron plays a similar role. IMHO, of course.
Saruman, I think, was not a flat evil character. He was clearly good at one point, and got too ambitious. Even then, he wasn't exactly evil -- he even tried to peacefully recruit Gandalf at one point, with the goal of going after Sauron, not becoming ruler of the world. He was kind of like Boromir -- he chose the wrong method to achive his goals, and ended up sinking deeper and deeper.
Max is interesting, but remember that he isn't actually all that complex.
The idea of much of the content in Max Payne is to build up a stereotypical character from film noir/comic books. It's not unenjoyable, just that it shouldn't be confused with being all that sophisticated.
The reason we see Max Payne as "fairly good" is that most video games are "fairly bad" when it comes to characters.
Remember what TV series frequently do with multiple episode story arcs? They have a quick introduction to where the story left off when they resume. It'd be pretty easy to include an option to "brief me on the story". When you combine that with a Neverwinter Nights-style journal for more complex games and the ability to "look" or "get info" on something or someone and see a description and things that you've learned about it, you can pretty much stop and go whenever you want.
Remember that lots of people read, say, a chapter of a book a night and still enjoy it, and a book can't do the kind of "quick recaps" that TV episodes do and video games *could* do.
Agreed.
The metric that video game plots and characters are judged by is wildly generous when compared to, say, movie equivalents. I hear people all the time saying that "Final Fantasy games have good plots" or "Final Fantasy games have good characters". Well, dammit, Final Fantasy really, really sucks in *both* areas compared to a good movie.
Some of that is because lots of games are intended to be played by children, and game authors (Nintendo produces a lot of these) find that it's easier to present folks with a very clear good guy and another very clear bad guy, maybe making the bad guy a bit goofy or suffering a number of character flaws and weaknesses.
Game budgets are big enough that Japanese-to-English translation and other technical obstacles of video games should not be a barrier anymore. We can expect *good* writing (good by *book* standards -- they have to write a lot less text and get a lot more money), engaging characters (same reason), and good plots. *Not* "good plot for a video game", which is what "good plot" has conventionally meant when discussing video games.
I remember when people went absolutely nuts when a video game actually included -- here it comes -- a *plot twist*! Awkward pauses in speech and unreadable expressions were also par for the course. Developers *can* do better than this, and *should*.