People always say that, but never explain "how it's a mess". This currently being written from XP x64, I've never seen what all of the fuss is about.
The only thing (generically speaking) that it seems to lack vs. Vista as far as an Actual Problem, is direct support for various AHCI drivers, so assuming you CAN install it (not that hard to figure out), it's hard to see how it could be considered 'a mess' otherwise, given that everything works, everything's compatible aside from a few (rare) outliers.
If you put middle aged women who constantly attend social functions in a padded cell, I bet they'd have 'withdrawl' too.
People crave interaction and attention. The internet can merely serve as a bridge to that for some people.
Some people of course take it rather far (constantly checking faceyspacey even when nothing happens), but the 'media' always links using any computer with 'internet addition' nowadays.
Grandmothers. Grandpas. Aunts. Uncles. The people who aren't close enough to you to care, and not with-it enough to do robust things like check, credit transfer, or god-forbid, an actual thoughtful gift. ^^;
Given the mention of PC...there's a good reason why it's #86 on PC (4 times lower than San Adreas), instead of #1.
The PC port was just unjustifiably buggy and lame, with Rockstar withholding fixes for months at a time.
Given that it's based on critic (not popular) review, you could even say that the 86 position is too damn good for it, especially since USERS give it a mere 4.6/10. http://www.metacritic.com/games/platforms/pc/grandtheftauto4
That, is a freaking trainwreck, especially given that it used particularly invasive form of SecuROM DRM which was the principle reason generally agreed upon (perhaps wayback has archives of the GTA4 forums just after release) for it performing so slow. http://www.pcgamefuntime.com/2008/12/grand-theft-auto-iv-drm-debacle/
You could throw a monster machine at it, and get 14-20FPS, even on low detail and low resolution.
If you point to how well received console versions were when somebody references the PC port, you clearly don't know what the hell you're talking about.
A decline in usage of IE6 doesn't mean XP is used any less, given that IE8 is available for XP as well.
Even according to the CLAIMED numbers, IE8 rose by 1.5%, Chrome rose 1%, and others mostly remained unchanged.
That would make IE8 the fastest growing browser, not Chrome. And a small increase on a web browser that nominally has 5%, tends to be within the realm of statistical noise, especially when all of the major browsers/OSes besides IE/Windows have seen fairly random multi-percent gains and drops over the last few years, with occasional hiccups in samples that seemed to mean something, but never panned out as a long term trend. And before somebody chimes in, it's a fixed "pot" of 100%. A growth of 1% on that scale is absolute, regardless of how much percentage you already have. 1.5% always means faster growth, even if you already have a high percentage.
This is a single snapshot, and one that's being mixed in with grease to turn the wheels. The numbers and claims are basically meaningless, and are the sort of "OMG!" story that turn up on reddit at least once a month.
There are even a few flubs, like saying that Firefox 3.6 was released mid-December, and they don't appear to include "IE8 compatibility mode" as part of IE8's numbers, despite it being the same browser, despite hitslink, what they're basing all of the data on, considering them different versions.
Chrome 4.0 and Firefox 3.6, released nearly the same time, both have very similar market share. 1.16% and 1.07%, respectively, which is impressive for new versions not even a week old.
On the non-version-specific list, IE has 62%, FF 24%, Chrome 5%, Safari 4%. As far as FF "declining", there just isn't that much room to grow if you can't wrestle it away from people who use Internet Explorer. Hitslinks' own TREND charts point out that FF has grown from 23% from March 2009 to 24% now. Hardly the loss these PC Magazine goons are pointing at now. Only IE suffered a meaningful loss, 68% to 62%, which is where all of the gains of EVERY other browser are coming out of.
Somehow, I think the real loser here is the integrity of tech journalism.
Given that he apparently believes that Earthquakes are the work of the devil...somehow, I don't see Pat Robertson advocating for proper physics education. Also, who the HELL watches MTV anymore? Seriously. I don't even think MTV4 still plays music videos (I remember when MTV2 had music videos! Gosh...and apparently I'm too young to've ever paid attention when MTV itself had music videos), and even VH1 stopped some years ago. Then again, I don't even know anyone who watches TV anymore, since the quality and level of intelligence displayed is so very...'lowest common denominator'.
If you're accusing an entire nation of 'sitting around all day watching MTV', you might just be projecting -a bit-.
That's why generally the extensions get approved after proving themselves stable/useful/coherent to the spec by the OpenGL ARB...and then still aren't supported by ATI.
ATI's OpenGL support has infamously been bad since the beginning, when I was getting backported support for OpenGL 1.5 on my Geforce 256, and it beat the Radeon 32DDR in support, stability, and driver support.
If Nvidia is the only one really putting backing force behind OpenGL (due to XGI's acquisition and S3's backing out of desktop graphics), that isn't the fault of the Khronous group, the ARB, the specification, or Nvidia. People always blame the predominant company if its competitors have simply made business decisions that made business analysts turn purple in confusion.
OpenGL 3.0 might have watered itself down to appease CAD, but all of the hooplah is largely years old, and irrelevant. It's also apparently largely ignored that 3.1 and 3.2 have been out for quite a while, and largely improved against 3.0 (and that Nvidia supported it rather quickly).
If you're still bitching about OpenGL 3.0 as an excuse, you're obviously in the DirectX-banner-waving-camp to begin with.
When it comes down to -my- personal opinion, for clarity, I think OpenGL is less obfuscated, and better focused (on graphics). The perk that you get software updates per-version, and don't require an entirely new revision of hardware in general to fully support new minor versions and major updates to functionality, API, and performance, is also something that rarely (if ever) happens with DirectX. Basically everything you need 'for OpenGL' is also supplied with the driver directly. How many people always complain in forums and troubleshooting tickets that "OMG, I haven't updated DirectX on this fresh install, how come it's not the latest version of special hack-on DLL that this game from 4 years ago needs, but is largely the same as all others??".
It's not officially supported, because it's a hacked implementation from the "Alky Project", and I doubt even the final source code they released after going under and not being able to stay funded.
It's not real DirectX 10. No hardware acceleration, just an API wrapper from an incomplete project that couldn't get funding, which noone has continued (despite source being out there). It doesn't work with most games, and can cause severe stability problems.
DirectX 10/11 for XP will not be possible unless Microsoft releases a service pack enabling WDDM drivers, which they've stated repeatedly they will not, will never do.
SeaMonkey (2.0) is based on the same components as Firefox (3.5), but the UI response is normal because it's extended from the original Mozilla Suite. Firefox, since its inception, has had single-thread, single-engine-for-entire-browser JS engine, which the UI (written in XUL) also uses. That typically leads to crap response, hanging if something doesn't expect that to happen.
SeaMonkey doesn't have that limitation. It doesn't have a 'faster JS engine' than FF (though Opera's is much slower), but responds much snappier, never 'hangs' if you have some JS taking up a lot of CPU time. It also generally uses less memory per-tab than FF, sadly enough, while being compatible with the most popular extensions (and there being extensions on addons.mozilla.com and xsidebar site making it act just like FF).
As perhaps the one caveat, though, certain Microsoft software (virtualearth/bing maps) refuse to recognize it or that it can handle SVG, throws a silent exception for which it -silently- mangles certain sites that use its service for mapping basic addresses and forms (cough, Pizza Hut, cough).
Would you rather not exist, or exist only to be harvested later?
Would you prefer animals that are very similar to your pets to've never been born, or treated and killed humanely to serve a purpose?
Etc and so on.
It's easy to say that the animals would've had better lives without this, but the fact is, they wouldn't have existed at all, given that their parents, and parents' parents, also were raised, lived, and died under similar conditions. These aren't wild animals plucked out of the forests and skinned still screeching.
It's easy to talk about cruelty about fur, or meat, but the human species as a whole is cruel not only to other animals, but to itself.
Given that the animals thus so far mentioned to be treated in such a manner, would also never have been alive if not for the very trade that demands they exist in such 'farms', it's more sensible to come up with a solution that humanely negates suffering, rather than make loud noises about how it shouldn't happen at all.
People will always buy crap, whether it's fur coats made out of an animal's pelt, or DVDs of "An Inconvenient Truth" made on former habitat of wolves.
The trick is, you need smarter, less greedy people, not more people arm-flailing about what people've been doing en masse ever since the species has existed. Arm-flailing fairly uselessly about the problem isn't doing anything to stop it, or educate out underlying social demand.
It just promotes the status quo of people wandering the streets saying "this is bad, mmkay?", and people responding with "pff, whatever". Religion, like veganism, is just another ideology with four-colour pamphlets in that respect.
If you want every fox, cat, dog, and any other random domesticated or farm-raised animal to survive with no purpose, then...all of the many millions can live with you, don't expect everyone else to house, feed, and clean up after them for you.
Ironically, the housing needs required for you and the precious un-wild animals would probably lead to the wholesale slaughter of many millions more WILD animals to procure and build on the land the necessary structures to keep them all in one place with means for food, shelter, and protection from the wild animals, who wouldn't hesitate to kill the farm-raised ones for food or being a perceived territorial threat.
The PDF itself says they outbred the fox population.
They're not left to inbreed amongst themselves indefinitely, or even apparently to any significant degree.
First, we designed
the mating system for our experimental fox
population to prevent it. Through outbreeding
with foxes from commercial fox farms and oth-
er standard methods, we have kept the in-
breeding coefficients for our fox population be-
tween 0.02 and 0.07. That means thatwhenever
a fox pup with a novel trait has been born into
the herd, the probability that it acquired the
trait through inbreeding (that is, by inheriting
both of its mutant genes from the same ances-
tor) has varied between only 2 and 7 percent.
It's silly to argue that they've been given significantly lower genetic diversity, given that they don't mate them significantly to each other, and that they effectively remove the least desirable (those that bite and show extreme fear from human contact) from the experiment. It's not a static and growing population of thousands of foxes, but a maintained breeding population of a largely fixed number (1996 had 700, 1999 only 100), from which only successive generations come, and numbers kept in careful control according to standard guidelines that cover this variety of experiment.
Given as of one of the last earth shattering maintainence releases (circa 2007), rxvt was about as useful as boiled spinach with nearly as many faults as the default, including deficiencies with copy-paste, extra-slow redrawing, and just as many problems with the various available curses libraries. I never said it was 'hard' to have it use rxvt, just that it wasn't especially worth it, by which I was particularly referring to 'the average person likely to install it'. I imagine some poor worker being told by their bosses that machines that can run fancy virtumimimal machines are too expensive for everyone to have, and needing some quick sed/awk/gcc light testing further than what the lightly maintained MSYS can provide.
Next time I'm forced to recompile some router firmware in linux-running virtual machine because of case sensitivity, source compatibility, utility issues, and generally quagmired times to execute any variety of long scripts, I'll be sure to remember that somebody mentioned how to add a shortcut to cygwin for rxvt. Not obvious at all, that. I'd prefer it if people were to read (and bother to notice context of reply) before hatchet-jobbing.
I never said it was the same, I was pointing out to the parent that both use the same broken cmd.exe terminal window by default.
Yes yes, and SUA was, and likely still is, given greater POSIX compliancy rating than Linux or FreeBSD by the government.
It's easy to say 'source compatible', but all that really means is that it hopefully conforms to some subset of relevant specifications of varying ages (such as UNIX98 or SUSv3). A great many (particularly smaller) applications specifically target Linux, which can bring equal trouble in trying to get them to compile or run on Cygwin, SUA, *BSD, Solaris, or what-have-you. Standards compliance doesn't seem to be looked on too favorably by terribly many programmers today.
rxvt just obviously isn't the default, and if somebody went through all of the trouble of installing and configuring X11 for Cygwin (which isn't even required for rxvt, or at least never used to be), wouldn't either A( running a VM, or B( actually installing a UNIX-like OS be the more natural option? UNIX != X11.
It's also lovely when people naturally assume that somebody saying 'not usually worth the trouble' means 'impossible for the writer to understand'.
Re:makes windows marginally bearable
on
Cygwin 1.7 Released
·
· Score: 1, Informative
Microsoft has had Interix, then Services For UNIX, and now Subsystem for Unix Applications originally since around 2000 but both SFU/SUA and Cygwin are pretty much just different shells on top of the limited cmd.exe window, unless you happen to use rxvt (which is usually not worth the trouble).
SFU 3.5 and now SUA for Server 2003 R2 and newer (including Vista and Windows 7) were free of charge, but it was comparable in most versions pay-or-not to MKS Toolkit and UWIN, rather than Cygwin and MingW (which are definitely more open-sourcey).
People recommend OpenWRT all the time, but when I tested with an Asus WL-520GU (virtually identical to the Linksys WRT54GL), the web UI and wireless drivers had serious CPU/IRQ problems on both OpenWRT and DD-WRT, even with wireless disabled, even with the very newest stable versions.
The Tomato firmware (some versions of which are modified/newer; it's not unmaintained on existing supported hardware), which runs on such devices, and has a much prettier (and functional) web interface has had no such issues or rebooting problems, regardless of the attached speed or wireless functionality. The 520gU also supports a USB port, which you can use with a hard drive or even crappy-cheap USB key to add swap and local storage. (The problems are solved on Tomato even without adding swap.) You can run bittorrent, TOR, and pretty much anything you want if it has a package (via optware), or you can cross-compile.
OpenWRT has native on-router compiler available through the same system, but the aforementioned problems seem to make it a moot point. OpenWRT seems like it might be better on stuff like the NetGear 3500L, which doubles most specs (480Mhz instead of 240, 64MB of ram instead of 16, 8MB flash instead of 4, n wireless instead of g).
There's also the point that Tomato is considerably easier to set up (and has inspiring bandwidth/classification tables/graphs to verify the QoS is actually having an effect) and manage unless you explicitly love the command line of iptables, or happen to enjoy Luci (which acts like a real pain on such hardware and has a habit of causing reboots, on both 2.4 and 2.6 kernels).
Mucking with page alignment and/or addressing would effectively prevent Nvidia/fglrx drivers from working (which is more or less why they don't work in HVM or the L4 microkernels, which implement Linux at a lower layer; they expect to be at specific addresses in a specific way, however you make it 'not that address', it doesn't work), nevermind wine, and it'd presumably be hard pressed to get a rootkit onto a well maintained Linux server in the first place, since nobody'd be running with root priveleges except a remote admin that logs in once in a while, with or without 'security modules' or Stack Smashing Protection on top to limit the scope and possibility of any intrusions or privilege escalation. So, this appears useless for desktops, useless for servers, what's left? It's good that money is being spent on research like this...
It would have to be generally assumed that everyone was referring to the 'big' games, and/or games otherwise that were of technical and artistic merit enough to warrant people still being interested in it over long term, and it having generic multiplay. Given that the number of games in that class is few to begin with (considerably less than one per year; completely nonexistant on consoles), it's also safe to assume that there are enough die-hard geeks to get around, and a fair majority will survive.
You're forgetting a few case examples: there are geeks running third party, wholly independent Microsoft Allegiance servers, and -nobody- played that game when it was live, nearly ten years ago. It effectively gets more players now, than Microsoft ever attracted with a fee.
WoW private servers are mostly dead when WoW goes down (if it happened today, mind), because everyone is complacent about the degree of bugs and incompleteness and the level of inflation. That's generally not the case with other private servers, such as EverQuest or Lineage 2, where if something isn't considered 'feature/game complete' at least to a specific expansion period, You have 100000 WoW private servers with 2 people each, vs a few hundred Lineage 2 servers with a few dozen people each, or a handful of EverQuest servers with a few hundred or thousand people each. Anecdotal, but...yeah. EverQuest private servers also kinda sucked in 2004 (five years after EQ's initial release, yes, EQEmu was around back then), Lineage 2 private servers were actually good in 2008 (five years after initial), but we're only five years out from the original World of Warcraft. Assuming anyone still cares, and project managers have dropped the idiocracy, then WoW private servers will have a shot at being pretty great in 2014, maybe. I know that 'everybody' wants them to hurry up, but in WoW's case vs. other games, it's significantly more a matter of project maintainers not caring about quality, while traditionally most of the projects for larger MMOs have strived for a large degree of perfection (and more than one alternative; competition breeds better output, there's only one type of WoW private server).
At the worst, Microsoft appears to be patenting some 'enhanced' variation of UAC (which doesn't ask for or present any of the things covered in the patent either, apparently).
This particular patent doesn't, as the parent points out, directly try to cover sudo, or Microsoft's own runas command, which -is- a direct ripoff of sudo.
As is apparently common in patents, Microsoft realized they wouldn't be granted protection (or likely even a review) anytime near when UAC originally came out, so they patented a leapfrog idea that they could possibly use years later, and wouldn't already be common public knowledge. The rules get murky at times, but even several Slashdot articles have covered this in the last year or two: you can't patent something that already exists, but you also can't patent something that you've already refused or delayed in patenting.
If Microsoft introduced UAC in Vista, but waited until after Windows 7, a second major product to use it or change it, to patent it, it'd dillute it enough (and WIndows 7's variation being minor) to be unpatentable by them or anyone else. MS leapfrogged the issue by having Vista UAC, and Windows 7, but neither did any of the major claims in the patent, which leaves them completely free to implement this in Windows 7.1 or 8.0, or whatever they decide to name the next version.
It's only 'fairly sucky' if you don't understand where the margins are and how they actually impact things. One of the reasons Firefox gets fairly notably bad interactive user performance is 'single JS thread used for the whole browser, including UI'. On the other hand, Chrome (on this machine at least) completes SunSpider in some 400ms, SM and FF around 800ms. Opera and most other browsers, which get very acceptable performance on 99% of things (production JS usage), are more in the 10000ms range. It's a series of specific microbenchmarks, remember, not the end-all be-all of JS benchmarks. At best, it can show where there are inefficiencies in code that CAN be called in tight loops (but more often is never called in such a way on real websites).
SeaMonkey essentially does (and has for years, albiet in a considereably less-useful form) fix all of the things that have been broken about Firefox's design since the beginning. Since it shares gecko rendering engine, and tracemonkey javascript engine (which is used in a less braindead manner in SM), it's not those features which are considered 'broken'. It is, essentially for people who preferred the better efficiency of classic Mozilla Suite, or are tired of FF's 'quirks', without having to give up all of the familiar and useful extensions (like Ad Block Plus, NoScript, CookieSafe, GreaseMonkey, etc).
Nowadays, finally, you can get similar functionality to FF (with tab mix plus and lots of other 'convenience'/safety extensions) on SM with compatible extensions, just with deterministic behavior and considerably less memory usage when using more than five tabs compared to the bloated manic-depressive fox.
Plausibly convincing AI doesn't strictly imply strong or particularly intelligent AI. Most games use inexpensive tricks to make the AI seem more competent than it actually is. Like telling it exactly where you are even when it can't see you, but applying extra fuzziness/inaccuracy to make it appear as if not cheating. But most modern games, including the original FEAR, are easy-mode, even on the hardest difficulty.
They cater to the lowest common denominator, when even most of the PC -and- console games from the late 1980s and early 1990s could be extremely difficult and time consuming for most people not using a game guide. But, most of the difficulty back then was game mechanics and novelty, not AI. Now that everything is decided by fairly comprehensive AI, CPU is cheap, the metric between difficulty and engaging gameplay has not kept up at all compared to the 'shiny graphics' everyone obsesses about. Graphics fidelity matters very little, so long as it matches itself and has a great level of consistency.
This obviously goes triply so for strategy games, it's even sadder when game developers can't figure out how to even make a half-way intelligent/communicative AI. Master of Orion 2 was perhaps the last game with particularly lucid diplomacy options where the AI responded in an intelligent way based on previous actions, current state, and future threats. Many modern strategy games have descended into Korean MMO-esque "grind fest". Capture 80 outposts, expand your borders, never have any internal defense, build up random economy/research improvements to make it go faster, rinse and repeat. The best expansionist always wins, even if they have (by far) the worst handicap.
Most games are an insult to a decent tactician, strategist, or someone merely coordinated and forward thinking. Lots of stuff to make it -look- convincing, but no depth, no real options, and very little (if any) challenge.
One recent criticism of game journalism pointed out (again) that most modern reviews (doubly so for 'mainstream' companies) are not significantly more than "this has the best graphics EVER, buy it!"
The litmus test against AI is whether it can actually adapt over time to your non-simple tactics and movements. Does it, can it figure out any of what you're doing, or does it sit there and fall for it -every time-? That's a pretty low bar, but a few games have managed to just scrape over it. Most don't even bother to try.
Welcome to the modern day shooting gallery, where things shoot back and take cover, but don't move much, and certainly never realize anything so simple as a basic flanking diversion, or that you have grenades and guns and can kill them.
So you're going to use an AMD IGP, and a more powerful Nvidia discrete addon card, but want to use the AMD IGP for rendering, and Nvidia only for PhysX? That's...great. Really.
The old adage goes: Speed, Reliability, Price. Pick any two.
Open source, performance, reliability, pick any two?
Nvidia and ATI don't generally have superior performance in IGPs either. They're IGPs, don't expect to get great performance on IGPs or lowest end cards (people still seem to brag about Geforce 7200GS, yagh) for gaming. That's obviously not what they're designed for, not even if they're brand new.
Nouveau has actually been going much faster, if you consider the actual age of the "radeon" driver (pre-2004 and still can barely draw a triangle on any card you can actually go out and still buy), vs. the actual age of the nouveau driver as a whole (2006 and is about the same state, but speeding up and merged into Mesa/DRM proper as well). You can 'vote with your dollars' by giving the money directly to an open source project which (usually) barely break even, if they actually do (few seem to), rather than corporations which earn billions of dollars per year and generally don't care what you do with your card.
It's kindof naive to spend money on AMD, and expect that the work will be done faster, rather than giving it to the people actually doing all of the hard work.
Personally, I'd rather have performance and reliability, so I'm content with binary drivers that can actually run things more complex than BZFlag and OpenArena at 20FPS. I'm also content to vote by Actual Development Time (ala contribution) on free projects, and saving money not superfluously yoyoing on video cards.
Each vendor makes promises they can't keep every ~2-2.5 years. If you haven't gone over the entire gamut in the last ten years for video cards which'd help you determine future trends in reliability/etc: in general concensus seems to be, AMD isn't reliable, isn't interested in fixing things, has been making the same promises since the Radeon 32DDR in 2001 (the most expensive consumer DirectX 7 part ever made), Nvidia seems to know what consumer wants, sometimes has hardware problems, sometimes does silly things to entice you that don't live up to hype, or are loud, power hungry behemoths that nobody can actually live with. Both are relatively a wash and rather even on the ACTUAL HARDWARE ITSELF, but the resulting differences, and problems that arise, are largely due to driver snafus each company makes that can persist for a year or more. Just, in my experience (I had a 'new' Radeon 3870 last year, mind), AMD's speeches about fixing things and changing graphics forever are significantly longer these days than the changelogs of their drivers, of which each release has fewer and fewer bugfixes, but fewer and fewer problems solved.
If you got band new shiny AMD card next week 'to support open source' (which it doesn't), and Nvidia open sourced all of their drivers first-party a day later, you'd be red in rage at AMD. Today, I can largely pick apart the Nvidia driver and 'fix' any problems necessary for API/ABI changes, which is generally not possible for AMD (which still lacks support for newer versions of the linux kernel, which has persistend since months before I pulled the 3870) . I'd love it if AMD made quality drivers that made my (probably less technically capable) 3870 zoom, I'd find a good home for it, but wishes and wants and future hopes are effectively irrelevant to the real world and what actually happens, unfortunately. Otherwise, I'd have working Nvidia (and ATI) drivers for FreeBSD amd64 and I could've stayed 'goodbye forever' to Windows for gaming.
Pick any which one and stick with it, period. You'll save a lot of money, and a lot of heartache. If you're indecisive and floundering over it every few years, it doesn't help anybody, not Nvidia, not AMD, certainly not open source developers, and least of all yourself.
All of which, of course, has nothing to do with rabblerousing about PhysX. People interested there should -really- get some technical background in why PhysX on ATI (and ATI primary Nvidia secondary) has never worked, and why TFA is poorly sourced only from an internet forum.
Umm. Should I even bother to point out how many things are wrong with what you said?
PhysX itself is used by developers. The card being capable of running it is fairly irrelevant. "Do you want shiny extra visual-only effects? Yes/No". PhysX GPU acceleration is fairly useless, and even many Nvidia users turn it off to save framerate.
You have a video card, which you presumably use for a game. If PhysX is used by the game...good for you, unless of course it uses PhysX software only (as the overwhelming majority oft PhysX licensees do). If it doesn't...or it's unsupported, have you somehow lost something? You're effectively putting to task the expectation that A( 'everything' uses PhysX, B( 'everything' must support PhysX.
You're effectively (and naively) blaming Nvidia for the lack of PhysX on AMD/ATI cards, when AMD itself hasn't been interested in it, hasn't developed its own comparable libraries, and let the Havok deal fall through. AMD has done, as far as I can tell, absolutely nothing to compete or interoperate in this area, not even encourage open source as do the work for them (as they're now doing for UNIX drivers).
The ultimate defense for libel is the truth, the ultimate defense for anti-competitive practices, is the competitor's own incompetence/inability to stay afloat. Nvidia had much (but not everything; NEC was the bigger factor there) to do with the demise of 3DFx, and absolutely nothing to do with the decline of AMD's competitive quality and their image in the eyes of the consumer. If more people bought AMD cards, then AMD would have a greater market share, logically. Nvidia hasn't sabotaged them in any way whatsoever, not even a teensy bit, nor do they hold the Lion's share of the GPU market.
Why not also blame Nvidia for not porting PhysX to Linux, MacOS, FreeBSD, and Solaris while you're at it? Come on. This is the real world, not 'every company does what you want before you want it' land.
The market for PhysX, is specifically developers, not you, the end user. Why are you so hell bent on getting a few extra graphical effects due to PhysX? That's what it's used for in most cases, and the most cases are...what, MAYBE 10 games released ever that directly support hardware PhysX from the GPU? And another 10 that supported the Aegia PPU, but not an Nvidia card?
Market share is also a non-issue for you, the end user. After all, do you get PhysX acceleration a Havok game either? No. Does it matter how much market share Havok or Bullet Physics have, if the game you like uses it? No.
You're...kinda wrong about MS and monopoly. They were ruled a monopoly for anti-competitive practices and various licensing deals they made to hurt the any competition. It has nothing to do with Apple 'not available on x86'. That's...very very strange. Apple backed IBM PowerPC, after they backed Motorola 68k. That it wasn't available on x86 was simply, they didn't want it to be. Just like they don't want it to be available for general non-Apple computers still today. Microsoft had nothing to do with that, and even periodically developed software for MacOS/PPC. They even developed a version of Windows NT 3.51 for PPC (which didn't hurt Apple at all).
Could AMD's loss of market share (and I seem to recall that Nvidia and ATI were fairly neck and neck in the Geforce 2/3 era) be due to their own problems which they have yet to resolve, despite ATI/AMD merger? Maybe possibly? If people don't trust you and your product to work particularly well, people will be less likely to put down $100+ merely for you to disappoint them. The door swings both ways.
Slashdot Mirror
People always say that, but never explain "how it's a mess".
This currently being written from XP x64, I've never seen what all of the fuss is about.
The only thing (generically speaking) that it seems to lack vs. Vista as far as an Actual Problem, is direct support for various AHCI drivers, so assuming you CAN install it (not that hard to figure out), it's hard to see how it could be considered 'a mess' otherwise, given that everything works, everything's compatible aside from a few (rare) outliers.
If you put middle aged women who constantly attend social functions in a padded cell, I bet they'd have 'withdrawl' too.
People crave interaction and attention. The internet can merely serve as a bridge to that for some people.
Some people of course take it rather far (constantly checking faceyspacey even when nothing happens), but the 'media' always links using any computer with 'internet addition' nowadays.
Grandmothers. Grandpas. Aunts. Uncles. The people who aren't close enough to you to care, and not with-it enough to do robust things like check, credit transfer, or god-forbid, an actual thoughtful gift. ^^;
Given the mention of PC...there's a good reason why it's #86 on PC (4 times lower than San Adreas), instead of #1.
The PC port was just unjustifiably buggy and lame, with Rockstar withholding fixes for months at a time.
Given that it's based on critic (not popular) review, you could even say that the 86 position is too damn good for it, especially since USERS give it a mere 4.6/10. http://www.metacritic.com/games/platforms/pc/grandtheftauto4
That, is a freaking trainwreck, especially given that it used particularly invasive form of SecuROM DRM which was the principle reason generally agreed upon (perhaps wayback has archives of the GTA4 forums just after release) for it performing so slow. http://www.pcgamefuntime.com/2008/12/grand-theft-auto-iv-drm-debacle/
You could throw a monster machine at it, and get 14-20FPS, even on low detail and low resolution.
If you point to how well received console versions were when somebody references the PC port, you clearly don't know what the hell you're talking about.
A decline in usage of IE6 doesn't mean XP is used any less, given that IE8 is available for XP as well.
Even according to the CLAIMED numbers, IE8 rose by 1.5%, Chrome rose 1%, and others mostly remained unchanged.
That would make IE8 the fastest growing browser, not Chrome. And a small increase on a web browser that nominally has 5%, tends to be within the realm of statistical noise, especially when all of the major browsers/OSes besides IE/Windows have seen fairly random multi-percent gains and drops over the last few years, with occasional hiccups in samples that seemed to mean something, but never panned out as a long term trend. And before somebody chimes in, it's a fixed "pot" of 100%. A growth of 1% on that scale is absolute, regardless of how much percentage you already have. 1.5% always means faster growth, even if you already have a high percentage.
This is a single snapshot, and one that's being mixed in with grease to turn the wheels. The numbers and claims are basically meaningless, and are the sort of "OMG!" story that turn up on reddit at least once a month.
There are even a few flubs, like saying that Firefox 3.6 was released mid-December, and they don't appear to include "IE8 compatibility mode" as part of IE8's numbers, despite it being the same browser, despite hitslink, what they're basing all of the data on, considering them different versions.
Chrome 4.0 and Firefox 3.6, released nearly the same time, both have very similar market share. 1.16% and 1.07%, respectively, which is impressive for new versions not even a week old.
On the non-version-specific list, IE has 62%, FF 24%, Chrome 5%, Safari 4%. As far as FF "declining", there just isn't that much room to grow if you can't wrestle it away from people who use Internet Explorer. Hitslinks' own TREND charts point out that FF has grown from 23% from March 2009 to 24% now. Hardly the loss these PC Magazine goons are pointing at now. Only IE suffered a meaningful loss, 68% to 62%, which is where all of the gains of EVERY other browser are coming out of.
Somehow, I think the real loser here is the integrity of tech journalism.
That sounds an awful lot like that old thing people haven't talked about since the 1990s...you know, a RENTAL. o.o
Given that he apparently believes that Earthquakes are the work of the devil...somehow, I don't see Pat Robertson advocating for proper physics education.
Also, who the HELL watches MTV anymore? Seriously. I don't even think MTV4 still plays music videos (I remember when MTV2 had music videos! Gosh...and apparently I'm too young to've ever paid attention when MTV itself had music videos), and even VH1 stopped some years ago.
Then again, I don't even know anyone who watches TV anymore, since the quality and level of intelligence displayed is so very...'lowest common denominator'.
If you're accusing an entire nation of 'sitting around all day watching MTV', you might just be projecting -a bit-.
That's why generally the extensions get approved after proving themselves stable/useful/coherent to the spec by the OpenGL ARB...and then still aren't supported by ATI.
ATI's OpenGL support has infamously been bad since the beginning, when I was getting backported support for OpenGL 1.5 on my Geforce 256, and it beat the Radeon 32DDR in support, stability, and driver support.
If Nvidia is the only one really putting backing force behind OpenGL (due to XGI's acquisition and S3's backing out of desktop graphics), that isn't the fault of the Khronous group, the ARB, the specification, or Nvidia. People always blame the predominant company if its competitors have simply made business decisions that made business analysts turn purple in confusion.
OpenGL 3.0 might have watered itself down to appease CAD, but all of the hooplah is largely years old, and irrelevant. It's also apparently largely ignored that 3.1 and 3.2 have been out for quite a while, and largely improved against 3.0 (and that Nvidia supported it rather quickly).
If you're still bitching about OpenGL 3.0 as an excuse, you're obviously in the DirectX-banner-waving-camp to begin with.
When it comes down to -my- personal opinion, for clarity, I think OpenGL is less obfuscated, and better focused (on graphics). The perk that you get software updates per-version, and don't require an entirely new revision of hardware in general to fully support new minor versions and major updates to functionality, API, and performance, is also something that rarely (if ever) happens with DirectX.
Basically everything you need 'for OpenGL' is also supplied with the driver directly. How many people always complain in forums and troubleshooting tickets that "OMG, I haven't updated DirectX on this fresh install, how come it's not the latest version of special hack-on DLL that this game from 4 years ago needs, but is largely the same as all others??".
It's not officially supported, because it's a hacked implementation from the "Alky Project", and I doubt even the final source code they released after going under and not being able to stay funded.
It's not real DirectX 10. No hardware acceleration, just an API wrapper from an incomplete project that couldn't get funding, which noone has continued (despite source being out there). It doesn't work with most games, and can cause severe stability problems.
DirectX 10/11 for XP will not be possible unless Microsoft releases a service pack enabling WDDM drivers, which they've stated repeatedly they will not, will never do.
SeaMonkey (2.0) is based on the same components as Firefox (3.5), but the UI response is normal because it's extended from the original Mozilla Suite. Firefox, since its inception, has had single-thread, single-engine-for-entire-browser JS engine, which the UI (written in XUL) also uses. That typically leads to crap response, hanging if something doesn't expect that to happen.
SeaMonkey doesn't have that limitation. It doesn't have a 'faster JS engine' than FF (though Opera's is much slower), but responds much snappier, never 'hangs' if you have some JS taking up a lot of CPU time. It also generally uses less memory per-tab than FF, sadly enough, while being compatible with the most popular extensions (and there being extensions on addons.mozilla.com and xsidebar site making it act just like FF).
As perhaps the one caveat, though, certain Microsoft software (virtualearth/bing maps) refuse to recognize it or that it can handle SVG, throws a silent exception for which it -silently- mangles certain sites that use its service for mapping basic addresses and forms (cough, Pizza Hut, cough).
Would you rather not exist, or exist only to be harvested later?
Would you prefer animals that are very similar to your pets to've never been born, or treated and killed humanely to serve a purpose?
Etc and so on.
It's easy to say that the animals would've had better lives without this, but the fact is, they wouldn't have existed at all, given that their parents, and parents' parents, also were raised, lived, and died under similar conditions. These aren't wild animals plucked out of the forests and skinned still screeching.
It's easy to talk about cruelty about fur, or meat, but the human species as a whole is cruel not only to other animals, but to itself.
Given that the animals thus so far mentioned to be treated in such a manner, would also never have been alive if not for the very trade that demands they exist in such 'farms', it's more sensible to come up with a solution that humanely negates suffering, rather than make loud noises about how it shouldn't happen at all.
People will always buy crap, whether it's fur coats made out of an animal's pelt, or DVDs of "An Inconvenient Truth" made on former habitat of wolves.
The trick is, you need smarter, less greedy people, not more people arm-flailing about what people've been doing en masse ever since the species has existed. Arm-flailing fairly uselessly about the problem isn't doing anything to stop it, or educate out underlying social demand.
It just promotes the status quo of people wandering the streets saying "this is bad, mmkay?", and people responding with "pff, whatever". Religion, like veganism, is just another ideology with four-colour pamphlets in that respect.
If you want every fox, cat, dog, and any other random domesticated or farm-raised animal to survive with no purpose, then...all of the many millions can live with you, don't expect everyone else to house, feed, and clean up after them for you.
Ironically, the housing needs required for you and the precious un-wild animals would probably lead to the wholesale slaughter of many millions more WILD animals to procure and build on the land the necessary structures to keep them all in one place with means for food, shelter, and protection from the wild animals, who wouldn't hesitate to kill the farm-raised ones for food or being a perceived territorial threat.
They're not left to inbreed amongst themselves indefinitely, or even apparently to any significant degree.
It's silly to argue that they've been given significantly lower genetic diversity, given that they don't mate them significantly to each other, and that they effectively remove the least desirable (those that bite and show extreme fear from human contact) from the experiment. It's not a static and growing population of thousands of foxes, but a maintained breeding population of a largely fixed number (1996 had 700, 1999 only 100), from which only successive generations come, and numbers kept in careful control according to standard guidelines that cover this variety of experiment.
Good day to you too.
Given as of one of the last earth shattering maintainence releases (circa 2007), rxvt was about as useful as boiled spinach with nearly as many faults as the default, including deficiencies with copy-paste, extra-slow redrawing, and just as many problems with the various available curses libraries.
I never said it was 'hard' to have it use rxvt, just that it wasn't especially worth it, by which I was particularly referring to 'the average person likely to install it'. I imagine some poor worker being told by their bosses that machines that can run fancy virtumimimal machines are too expensive for everyone to have, and needing some quick sed/awk/gcc light testing further than what the lightly maintained MSYS can provide.
Next time I'm forced to recompile some router firmware in linux-running virtual machine because of case sensitivity, source compatibility, utility issues, and generally quagmired times to execute any variety of long scripts, I'll be sure to remember that somebody mentioned how to add a shortcut to cygwin for rxvt. Not obvious at all, that.
I'd prefer it if people were to read (and bother to notice context of reply) before hatchet-jobbing.
I never said it was the same, I was pointing out to the parent that both use the same broken cmd.exe terminal window by default.
Yes yes, and SUA was, and likely still is, given greater POSIX compliancy rating than Linux or FreeBSD by the government.
It's easy to say 'source compatible', but all that really means is that it hopefully conforms to some subset of relevant specifications of varying ages (such as UNIX98 or SUSv3).
A great many (particularly smaller) applications specifically target Linux, which can bring equal trouble in trying to get them to compile or run on Cygwin, SUA, *BSD, Solaris, or what-have-you. Standards compliance doesn't seem to be looked on too favorably by terribly many programmers today.
rxvt just obviously isn't the default, and if somebody went through all of the trouble of installing and configuring X11 for Cygwin (which isn't even required for rxvt, or at least never used to be), wouldn't either A( running a VM, or B( actually installing a UNIX-like OS be the more natural option? UNIX != X11.
It's also lovely when people naturally assume that somebody saying 'not usually worth the trouble' means 'impossible for the writer to understand'.
Microsoft has had Interix, then Services For UNIX, and now Subsystem for Unix Applications originally since around 2000 but both SFU/SUA and Cygwin are pretty much just different shells on top of the limited cmd.exe window, unless you happen to use rxvt (which is usually not worth the trouble).
SFU 3.5 and now SUA for Server 2003 R2 and newer (including Vista and Windows 7) were free of charge, but it was comparable in most versions pay-or-not to MKS Toolkit and UWIN, rather than Cygwin and MingW (which are definitely more open-sourcey).
People recommend OpenWRT all the time, but when I tested with an Asus WL-520GU (virtually identical to the Linksys WRT54GL), the web UI and wireless drivers had serious CPU/IRQ problems on both OpenWRT and DD-WRT, even with wireless disabled, even with the very newest stable versions.
The Tomato firmware (some versions of which are modified/newer; it's not unmaintained on existing supported hardware), which runs on such devices, and has a much prettier (and functional) web interface has had no such issues or rebooting problems, regardless of the attached speed or wireless functionality. The 520gU also supports a USB port, which you can use with a hard drive or even crappy-cheap USB key to add swap and local storage. (The problems are solved on Tomato even without adding swap.) You can run bittorrent, TOR, and pretty much anything you want if it has a package (via optware), or you can cross-compile.
OpenWRT has native on-router compiler available through the same system, but the aforementioned problems seem to make it a moot point. OpenWRT seems like it might be better on stuff like the NetGear 3500L, which doubles most specs (480Mhz instead of 240, 64MB of ram instead of 16, 8MB flash instead of 4, n wireless instead of g).
There's also the point that Tomato is considerably easier to set up (and has inspiring bandwidth/classification tables/graphs to verify the QoS is actually having an effect) and manage unless you explicitly love the command line of iptables, or happen to enjoy Luci (which acts like a real pain on such hardware and has a habit of causing reboots, on both 2.4 and 2.6 kernels).
Mucking with page alignment and/or addressing would effectively prevent Nvidia/fglrx drivers from working (which is more or less why they don't work in HVM or the L4 microkernels, which implement Linux at a lower layer; they expect to be at specific addresses in a specific way, however you make it 'not that address', it doesn't work), nevermind wine, and it'd presumably be hard pressed to get a rootkit onto a well maintained Linux server in the first place, since nobody'd be running with root priveleges except a remote admin that logs in once in a while, with or without 'security modules' or Stack Smashing Protection on top to limit the scope and possibility of any intrusions or privilege escalation.
So, this appears useless for desktops, useless for servers, what's left? It's good that money is being spent on research like this...
It would have to be generally assumed that everyone was referring to the 'big' games, and/or games otherwise that were of technical and artistic merit enough to warrant people still being interested in it over long term, and it having generic multiplay. Given that the number of games in that class is few to begin with (considerably less than one per year; completely nonexistant on consoles), it's also safe to assume that there are enough die-hard geeks to get around, and a fair majority will survive.
You're forgetting a few case examples: there are geeks running third party, wholly independent Microsoft Allegiance servers, and -nobody- played that game when it was live, nearly ten years ago. It effectively gets more players now, than Microsoft ever attracted with a fee.
WoW private servers are mostly dead when WoW goes down (if it happened today, mind), because everyone is complacent about the degree of bugs and incompleteness and the level of inflation. That's generally not the case with other private servers, such as EverQuest or Lineage 2, where if something isn't considered 'feature/game complete' at least to a specific expansion period,
You have 100000 WoW private servers with 2 people each, vs a few hundred Lineage 2 servers with a few dozen people each, or a handful of EverQuest servers with a few hundred or thousand people each. Anecdotal, but...yeah.
EverQuest private servers also kinda sucked in 2004 (five years after EQ's initial release, yes, EQEmu was around back then), Lineage 2 private servers were actually good in 2008 (five years after initial), but we're only five years out from the original World of Warcraft. Assuming anyone still cares, and project managers have dropped the idiocracy, then WoW private servers will have a shot at being pretty great in 2014, maybe. I know that 'everybody' wants them to hurry up, but in WoW's case vs. other games, it's significantly more a matter of project maintainers not caring about quality, while traditionally most of the projects for larger MMOs have strived for a large degree of perfection (and more than one alternative; competition breeds better output, there's only one type of WoW private server).
At the worst, Microsoft appears to be patenting some 'enhanced' variation of UAC (which doesn't ask for or present any of the things covered in the patent either, apparently).
This particular patent doesn't, as the parent points out, directly try to cover sudo, or Microsoft's own runas command, which -is- a direct ripoff of sudo.
As is apparently common in patents, Microsoft realized they wouldn't be granted protection (or likely even a review) anytime near when UAC originally came out, so they patented a leapfrog idea that they could possibly use years later, and wouldn't already be common public knowledge. The rules get murky at times, but even several Slashdot articles have covered this in the last year or two: you can't patent something that already exists, but you also can't patent something that you've already refused or delayed in patenting.
If Microsoft introduced UAC in Vista, but waited until after Windows 7, a second major product to use it or change it, to patent it, it'd dillute it enough (and WIndows 7's variation being minor) to be unpatentable by them or anyone else. MS leapfrogged the issue by having Vista UAC, and Windows 7, but neither did any of the major claims in the patent, which leaves them completely free to implement this in Windows 7.1 or 8.0, or whatever they decide to name the next version.
It's only 'fairly sucky' if you don't understand where the margins are and how they actually impact things.
One of the reasons Firefox gets fairly notably bad interactive user performance is 'single JS thread used for the whole browser, including UI'.
On the other hand, Chrome (on this machine at least) completes SunSpider in some 400ms, SM and FF around 800ms. Opera and most other browsers, which get very acceptable performance on 99% of things (production JS usage), are more in the 10000ms range. It's a series of specific microbenchmarks, remember, not the end-all be-all of JS benchmarks. At best, it can show where there are inefficiencies in code that CAN be called in tight loops (but more often is never called in such a way on real websites).
SeaMonkey essentially does (and has for years, albiet in a considereably less-useful form) fix all of the things that have been broken about Firefox's design since the beginning. Since it shares gecko rendering engine, and tracemonkey javascript engine (which is used in a less braindead manner in SM), it's not those features which are considered 'broken'.
It is, essentially for people who preferred the better efficiency of classic Mozilla Suite, or are tired of FF's 'quirks', without having to give up all of the familiar and useful extensions (like Ad Block Plus, NoScript, CookieSafe, GreaseMonkey, etc).
Nowadays, finally, you can get similar functionality to FF (with tab mix plus and lots of other 'convenience'/safety extensions) on SM with compatible extensions, just with deterministic behavior and considerably less memory usage when using more than five tabs compared to the bloated manic-depressive fox.
Plausibly convincing AI doesn't strictly imply strong or particularly intelligent AI. Most games use inexpensive tricks to make the AI seem more competent than it actually is. Like telling it exactly where you are even when it can't see you, but applying extra fuzziness/inaccuracy to make it appear as if not cheating. But most modern games, including the original FEAR, are easy-mode, even on the hardest difficulty.
They cater to the lowest common denominator, when even most of the PC -and- console games from the late 1980s and early 1990s could be extremely difficult and time consuming for most people not using a game guide. But, most of the difficulty back then was game mechanics and novelty, not AI. Now that everything is decided by fairly comprehensive AI, CPU is cheap, the metric between difficulty and engaging gameplay has not kept up at all compared to the 'shiny graphics' everyone obsesses about. Graphics fidelity matters very little, so long as it matches itself and has a great level of consistency.
This obviously goes triply so for strategy games, it's even sadder when game developers can't figure out how to even make a half-way intelligent/communicative AI. Master of Orion 2 was perhaps the last game with particularly lucid diplomacy options where the AI responded in an intelligent way based on previous actions, current state, and future threats. Many modern strategy games have descended into Korean MMO-esque "grind fest". Capture 80 outposts, expand your borders, never have any internal defense, build up random economy/research improvements to make it go faster, rinse and repeat. The best expansionist always wins, even if they have (by far) the worst handicap.
Most games are an insult to a decent tactician, strategist, or someone merely coordinated and forward thinking. Lots of stuff to make it -look- convincing, but no depth, no real options, and very little (if any) challenge.
One recent criticism of game journalism pointed out (again) that most modern reviews (doubly so for 'mainstream' companies) are not significantly more than "this has the best graphics EVER, buy it!"
The litmus test against AI is whether it can actually adapt over time to your non-simple tactics and movements. Does it, can it figure out any of what you're doing, or does it sit there and fall for it -every time-? That's a pretty low bar, but a few games have managed to just scrape over it. Most don't even bother to try.
Welcome to the modern day shooting gallery, where things shoot back and take cover, but don't move much, and certainly never realize anything so simple as a basic flanking diversion, or that you have grenades and guns and can kill them.
So you're going to use an AMD IGP, and a more powerful Nvidia discrete addon card, but want to use the AMD IGP for rendering, and Nvidia only for PhysX?
That's...great. Really.
The old adage goes: Speed, Reliability, Price. Pick any two.
Open source, performance, reliability, pick any two?
Nvidia and ATI don't generally have superior performance in IGPs either. They're IGPs, don't expect to get great performance on IGPs or lowest end cards (people still seem to brag about Geforce 7200GS, yagh) for gaming. That's obviously not what they're designed for, not even if they're brand new.
Nouveau has actually been going much faster, if you consider the actual age of the "radeon" driver (pre-2004 and still can barely draw a triangle on any card you can actually go out and still buy), vs. the actual age of the nouveau driver as a whole (2006 and is about the same state, but speeding up and merged into Mesa/DRM proper as well).
You can 'vote with your dollars' by giving the money directly to an open source project which (usually) barely break even, if they actually do (few seem to), rather than corporations which earn billions of dollars per year and generally don't care what you do with your card.
It's kindof naive to spend money on AMD, and expect that the work will be done faster, rather than giving it to the people actually doing all of the hard work.
Personally, I'd rather have performance and reliability, so I'm content with binary drivers that can actually run things more complex than BZFlag and OpenArena at 20FPS.
I'm also content to vote by Actual Development Time (ala contribution) on free projects, and saving money not superfluously yoyoing on video cards.
Each vendor makes promises they can't keep every ~2-2.5 years. If you haven't gone over the entire gamut in the last ten years for video cards which'd help you determine future trends in reliability/etc: in general concensus seems to be, AMD isn't reliable, isn't interested in fixing things, has been making the same promises since the Radeon 32DDR in 2001 (the most expensive consumer DirectX 7 part ever made), Nvidia seems to know what consumer wants, sometimes has hardware problems, sometimes does silly things to entice you that don't live up to hype, or are loud, power hungry behemoths that nobody can actually live with. Both are relatively a wash and rather even on the ACTUAL HARDWARE ITSELF, but the resulting differences, and problems that arise, are largely due to driver snafus each company makes that can persist for a year or more. Just, in my experience (I had a 'new' Radeon 3870 last year, mind), AMD's speeches about fixing things and changing graphics forever are significantly longer these days than the changelogs of their drivers, of which each release has fewer and fewer bugfixes, but fewer and fewer problems solved.
If you got band new shiny AMD card next week 'to support open source' (which it doesn't), and Nvidia open sourced all of their drivers first-party a day later, you'd be red in rage at AMD. Today, I can largely pick apart the Nvidia driver and 'fix' any problems necessary for API/ABI changes, which is generally not possible for AMD (which still lacks support for newer versions of the linux kernel, which has persistend since months before I pulled the 3870) . I'd love it if AMD made quality drivers that made my (probably less technically capable) 3870 zoom, I'd find a good home for it, but wishes and wants and future hopes are effectively irrelevant to the real world and what actually happens, unfortunately. Otherwise, I'd have working Nvidia (and ATI) drivers for FreeBSD amd64 and I could've stayed 'goodbye forever' to Windows for gaming.
Pick any which one and stick with it, period. You'll save a lot of money, and a lot of heartache. If you're indecisive and floundering over it every few years, it doesn't help anybody, not Nvidia, not AMD, certainly not open source developers, and least of all yourself.
All of which, of course, has nothing to do with rabblerousing about PhysX. People interested there should -really- get some technical background in why PhysX on ATI (and ATI primary Nvidia secondary) has never worked, and why TFA is poorly sourced only from an internet forum.
Umm. Should I even bother to point out how many things are wrong with what you said?
PhysX itself is used by developers. The card being capable of running it is fairly irrelevant. "Do you want shiny extra visual-only effects? Yes/No". PhysX GPU acceleration is fairly useless, and even many Nvidia users turn it off to save framerate.
You have a video card, which you presumably use for a game. If PhysX is used by the game...good for you, unless of course it uses PhysX software only (as the overwhelming majority oft PhysX licensees do). If it doesn't...or it's unsupported, have you somehow lost something?
You're effectively putting to task the expectation that A( 'everything' uses PhysX, B( 'everything' must support PhysX.
You're effectively (and naively) blaming Nvidia for the lack of PhysX on AMD/ATI cards, when AMD itself hasn't been interested in it, hasn't developed its own comparable libraries, and let the Havok deal fall through. AMD has done, as far as I can tell, absolutely nothing to compete or interoperate in this area, not even encourage open source as do the work for them (as they're now doing for UNIX drivers).
The ultimate defense for libel is the truth, the ultimate defense for anti-competitive practices, is the competitor's own incompetence/inability to stay afloat. Nvidia had much (but not everything; NEC was the bigger factor there) to do with the demise of 3DFx, and absolutely nothing to do with the decline of AMD's competitive quality and their image in the eyes of the consumer. If more people bought AMD cards, then AMD would have a greater market share, logically. Nvidia hasn't sabotaged them in any way whatsoever, not even a teensy bit, nor do they hold the Lion's share of the GPU market.
Why not also blame Nvidia for not porting PhysX to Linux, MacOS, FreeBSD, and Solaris while you're at it? Come on. This is the real world, not 'every company does what you want before you want it' land.
The market for PhysX, is specifically developers, not you, the end user. Why are you so hell bent on getting a few extra graphical effects due to PhysX? That's what it's used for in most cases, and the most cases are...what, MAYBE 10 games released ever that directly support hardware PhysX from the GPU? And another 10 that supported the Aegia PPU, but not an Nvidia card?
Market share is also a non-issue for you, the end user. After all, do you get PhysX acceleration a Havok game either? No. Does it matter how much market share Havok or Bullet Physics have, if the game you like uses it? No.
You're...kinda wrong about MS and monopoly. They were ruled a monopoly for anti-competitive practices and various licensing deals they made to hurt the any competition. It has nothing to do with Apple 'not available on x86'. That's...very very strange. Apple backed IBM PowerPC, after they backed Motorola 68k. That it wasn't available on x86 was simply, they didn't want it to be. Just like they don't want it to be available for general non-Apple computers still today. Microsoft had nothing to do with that, and even periodically developed software for MacOS/PPC. They even developed a version of Windows NT 3.51 for PPC (which didn't hurt Apple at all).
Nvidia is far from a monopoly in the graphics market, with a mere 29% of the market and AMD 17%. Intel dominates with 44%. At least according to JPR, whom I'm sure you distrust as well: http://jonpeddie.com/press-releases/details/amd-soars-in-q209-intel-and-nvidia-also-show-great-gains/
Could AMD's loss of market share (and I seem to recall that Nvidia and ATI were fairly neck and neck in the Geforce 2/3 era) be due to their own problems which they have yet to resolve, despite ATI/AMD merger? Maybe possibly? If people don't trust you and your product to work particularly well, people will be less likely to put down $100+ merely for you to disappoint them. The door swings both ways.