Initially I moderated this comment "Insightful", but decided instead to explain myself. He asked a perfectly legitimate question and got modded down twice. Come on, people - use your noggins instead of proving him right.
Sigh. And now it's at -1.
I think it's fascinating to watch kneejerk moderators at work: I don't understand why certain people get so fanatical about products (and this applies to many products, not just computers) so I ask if anybody else does and get modded as flamebait. Yes, it's clearly flamebait, as in an Apple topic anything that doesn't praise his Jobsness is obviously inflammatory.
Jesus guys, that's my first -1 for over 8 months now, I do not post flamebait. I might have guessed my next would come from Apple fanboy moderators.
Re:AltaVista vs. Google: speed and relevance shoot
on
Altavista Renewed
·
· Score: 3, Informative
Actually, minority OS users do own the net:) Try doing a google search (and this works on altavista too) for Wine - the first link goes to WineHQ for the Windows emulator.
I'd love to gain some insight into the psychology behind people who become fanatical over products. What exactly motivates people to do all these crazy things in the name of a piece of electronics?
Now please don't give me any BS about how it's because the Mac is "so high quality" and "a classy piece of gear" as a Machead once tried to convince me, you can get high quality cars, stereo systems, hell even high quality food, but you don't see people build shrines to the Porsche do you? Or maybe you do.....
If anything I can kind of understand Linux freaks, as it's more a movement (vive la revolution and all that) than a product. But the Mac isn't anything other than some heavily marketed electronics. What motivates these people?
The W3C has given us JavaScript+DOM+CSS+..., but it's way too complicated for the vanishingly small amount of functionality, and nobody has managed to implement it correctly; in fact, I doubt nobody knows what a correct implementation would even mean.
Huh? JavaScript is the Mozilla implementation of ECMAScript, a standard (not W3C) invented by Netscape. The DOM was also a Netscape idea, now standardized. CSS was originally proposed and largely designed by a guy from Opera. There are quite a few implementations out there actually, the idea that W3C technologies are too large to implement is crazy. Look at Mozilla, Amaya, even Konqueror is getting there now.....
The W3C should have stopped with a full specification of HTML. Anything they have been doing beyond that has been doing more harm than good. The web succeeded because HTML was simple.
Yes, and now it's ubiquitous do you really think we need to keep it simple? Being simple was great when the web was small, it let it grow very quickly. Why should we keep it simple now? Just for the sake of it? I'd rather have power. If that means there are only 3 or 4 quality implementations as opposed to 20, then so be it.
The world is not a simple place, and the things we want to do with the web nowadays aren't simple either. If you want simplicity then feel free to write a web browser that only understands a subset of the standards, they are layered so people can do this. Just bear in mind that it won't be useful for browsing the web, because a lot of people like powerful technologies and use them.
What good is a standard if you never hold anyone's feet to the fire if they don't support it? If developers never have any incentive to actually get it right? If the standards are so vague that it allows for interpretations that can be so drastically different that the standard becomes useless?
You have to have standards. The W3C are the people who are widely recognized as being the technical lead for the net. Now they don't make law, quite right, but if there was no W3C then Microsoft really WOULD own the web: as it is, we can and do take them to task when they break the rules. They can ignore us of course, yet whaddaya know but IE6 supports DOM/CSS Level 1. Not a particularly impressive achievement, but it's a start.
The standards are actually very precise, which is one reason they are seens as being very large. There is hardly any room for interpretation in stuff like the DOM, CSS, XML etc. Of course, sometimes when the internal architecture of IE mandates it Microsoft simply ignore things, the mime-type issue being a good example, but also the fact that you have to specify node.className = "class" to set the style on a new element, as opposed to setting the class attribute (which works fine in Mozilla). Why? Because (according to an MS developer) internally the MS dom is based on object model attributes, so that's what you have to set.
Has any company yet written a complete CSS1 implementation? A complete working version of DOM0? Yet here we are toiling away on XHTML and CSS3(!) and DOM Level 2. And they don't even seem to give a rat's ass if anyone actually follows the rules.
[sigh] Yes. Mozilla supports DOM and CSS Level 2 and they have partial support for Level 3 now. Level 0 is the term used to refer to the pre-standardized technologies, it doesn't actually exist as a standard so EVERY browser that can script web pages has a level zero DOM. It should be noted that TBL himself has stepped in on occasion to tick off Microsoft about stuff like browser blocks, bad HTML etc.
From what I hear about CSS3, it's going to be such a massive specification that no company (save Microsoft, if they actually gave a damn) would possibly be able to implement it.
Then you hear wrong.
In the meantime we see developers actually building websites entirely out of Flash because there's one reference implementation (one version, period) and it just works. Is that the future we want?
Developers do not build web pages out of flash. Marketing departments do. Luckily most web pages are not built by marketing.
It's time to hold these clowns accountable. Make them do some real work: make them create a working version of their spec.
Poor troll. The W3C already implement all their standards, go to w3.org and download Amaya. Nobody uses it for actually browsing the web, but there it is, proof that an actually very small organization with very few coders can implement their standards.
Of course, if ISPs really wanted to put a long term solution for this in place, they'd get off their backsides and put IPv6 into place. All modern operating systems support it, and with IPv6 comes decent IP Multicast.
This whole situation has been caused simply because a network stream is a 1:1 thing. If you make it a 1:many connection then suddenly you have far more efficient use of bandwidth, because if you're an ISP and 100 users are downloading a file, you only have to receive it once on your incoming link, instead of 100 times.
Of course, you'd still have P2P file sharing, the reason that people use Kazaa is not because it's the best way of moving information around (it's not), it's because it's anonymous and you can't be taken down for it. Safety in numbers, safety in anonymity. If there were suddenly large pirate music servers transmitting albums on rotation via multicast 24/7 they would be much easier targets.
Multicast has lots of other legal uses of course, that's what I'd want them for. But I can see that it'd help solve this situation. So come on ISPs, where are the v6 routers?
IIRC that is exactly what Kazaa does. It's impossible to block on a port by port basis as it simply switches to different ports, and all the traffic on the FastTrack network is encrypted. You can still get to it though, despite being encrypted FastTrack traffic has recognizable headers and traffic signatures. Often places like universities use packet shapers to try and eliminate it. Unfortunately this has the effect of often massively increasing latency, killing multiplayer games. It can also knock out legitimate traffic.
The simplest solution is simply to place upload caps. You wouldn't eliminate all p2p users, but that's OK, you just want to hit the guys who constantly upload stuff 24/7.
The great irony would be that Microsoft, who came late to the browser game, could then sue Netscape/Mozilla out of existence
Unlikely. As it's already been pointed out above, if MS was to be sued under this patent, they'd simply pull ActiveX from IE (if it held, which it wouldn't). The world would continue to go around, albiet the girl in my office would wonder why she can't play scrabble on MSN Games anymore.
If Microsoft got these patents and sued Mozilla, what would happen? Jack all is my answer. Who would they sue? The Mozilla organization? Can they even be held legally responsible for Mozilla? And if they could, they still couldn't destroy Mozilla, it's an open source project, we all have it, and it would continue to be worked on.
Years ago MS was talking about using patents as a weapon against free software - but I've not seen anything yet. Have you?
Ubiquitous within Linux, what did you think? I'm assuming a Linux based network.
Oh, and when will X get audio support? I can play back movies with sound through Terminal Server, even over my DSL. So maybe I don't have as great an experience with D3D via TS, but I've never been able to play Quake over X over DSL on X.
Huh? You're confused. X is a graphics protocol, audio is outside its remit. If you want network transparent audio you can setup aRts or eSound. It'd be nice to have it integrated with the DISPLAY variable or something, but that's pretty trivial to do.
As for Quake, I have no idea how you managed to play a fullscreen 3D game over terminal services (which is based on compressed bitmaps) on DSL, it must have been a very low latency line. You can play games over X though, I do know that, as it sends the opengl calls themselves through the connection.
On a related note, where are all those great applications for X on Linux? Things like Adobe Photoshop or Paint Shop Pro or Avid or Finale or IE. Can you name me one app for Linux that doesn't have a port or equivalent on Windows?
Tip: when trolling, don't pick bad examples. Adobe Photoshop/PSP = The gimp. Yes, that exists on Windows too, because somebody ported it. That's software freedom working for you. IE = Mozilla. Avid?? dunno what that is, ditto for Finale. There are almost certainly equivalents for joe-level apps, if there are specialists things you need you can use Wine. How did we get onto this anyway, is this just some kind of general anti-linux rant? I was talking about applying X to tablet pcs.
Your thumbnails example is a terrible one. You're Linux geek friend was probably in the middle of doing something wierd to his machine, if I wanted to view a CD of images (i'm assuming they're in a standard format, not one MS pulled out of its ass) then I just put the CD in the drive and Nautilus or Konquerer will open and compile thumbnails for me. I know from experience they do this very quickly indeed. Maybe your friend eschews the easy way or something, I dunno.
I really have to laugh at anyone who thinks they can be better than MS at integrating technologies around Linux and competing on the Tablet platform before Microsoft eats it all.
Laugh then. We'll laugh last, guaranteed. This is such a pathetic pro-MS rant it's untrue. Why am I bothering to reply? Oh, I know, it's raining and I don't want to get wet by going home just yet.
If Microsoft has never, ever let you down then by definition you are new to the computing scene. Having friends who can get you cheap copies of Windows is pathetic, and that last comment is ludicrous - I got redhat for free! And I didn't break the law or call favours. It's more stable than Windows, virus free, lots of good support etc etc etc.
Jesus. I give up. I hope you weren't serious, if so then you should really do some investigations instead of trolling slashdot, you might learn something.
Mainly because these things are actually very hard. For instance, package management (software installation) - dragging stuff from a CD or the Net sounds easy, and it is, but you've just removed a lot of power to get that simplicity. For one, what happens to dependancies? I guess they could be auto-located and installed as the system "copied" the file or something, but what if one can't be installed for whatever reason? Copy failed? I dunno. Interesting to work on however. Note the way OS X does it is imho dumb, you don't even get a chance to do dependancies because you are in fact copying a directory. Having apps represented in the filing system is a good idea, having them actually be atomic FS objects with todays filing system technology is nuts.
Live queries - coming, at least in ReiserFS and apparently Longhorn. I've not used BeOS but I find simply organizing my files well into directories means I never want to do searches anyway.
Version control in filing systems isn't a new idea, but again, the devil is in the details. What happens if you run out of disk space (i for instance have only a 10gig disk and it's already full). Do you start randomly killing past revisions? How do you store the different versions? Plain text is easy, you just diff it, but what about images? Store a whole new copy of the image each time perhaps, because as far as I know there is no image diffing technology.
Perhaps a better idea is to store transactions. Any program that supports multi-level undo probably internally is structured around transactions, ie objects that represent an alteration to the document. A document could be split up into 2 parts: the data itself and a transactions queue. When you start a new file, the file stores your actions, rather than the actual document state, and loading it involves rebuilding the document from the transactions. This would allow theoretically infinite undo, limited only by disk space. Of course rerunning actions would take ages, so you could have a throwaway cache of the current state of the document.
Erm, what else? The RISC OS style filepickers are available for use in your apps via the rox library, and yes, they do rock. You can do this in Windows at any rate by opening up an explorer window, finding the document and dragging it onto the window. It doesn't work very well however. It would be relatively easy to add XDS support to the GTK file picker (it's being rewritten by hadess i think) so if you want you can use drag and drop save/load.
Application loading/unloading - there's no real reason why we should keep this, he's quite right, and really we don't have it anymore. When I want to browse the web, I click on the web browser button in my panel, and up pops a Mozilla window. If Moz isn't loaded it takes a few more seconds than normal, but so what? It's multi-instance. When I logout, my apps persist themselves to the session manager so they are back again when I login. I don't think much about loading/unloading apps anymore, that's just a dumb Mac thing - one they could kill so easily, but didn't.
What MPT is really ranting about here is not what he thinks he is. Computers have history, and with history comes inertia. Support linux man! You can add most of what you've been talking about to apps pretty easily when all the code is open. I can think of ways of doing all those things with slight modifications to toolkits and apps.
Oh boy, you hit that on the nail. My gripes with soap are extensive. For one, the way of passing XML as parameters or return values appears to have been removed in SOAP 1.1 - there used to be a literalXML attribute you could use to signal that a method was returning actual XML as opposed to a data structure. Where is it now? I dunno, and as the W3C helpfully don't appear to list the preceding versions of the spec on their site (i feel they must be there, but never found them) I can't find out if I'm simply going crazy or they really did remove it.
When a frustrated person on the list asked the same questions I'd been having, he was told there seemed to be an informal standard of using an Apache Axis namespace for XML now! I mean WTF? But that's only the beginning.
Where is objects-by-reference? Oh, I know, that'd be in the Microsoft.NET Remoting extensions. Why not in the standard itself? Because MS have been blocking it on the grounds of "added bloat". I'd say objects-by-reference are absolutely critical to any modern RPC protocol and it doesn't have to be complex either, but it's not in there. Simple Object Access Protocol my backside.
And then of course the patents. Considering we've been doing RPC for over a decade, and SOAP is merely RPC done in XML (and done worse) how in the lords name can there be patents on it? Prior art trips you up at the street corner there's so much. And WSDL? What a POS. That is a classic example of something that's been abstracted so much it's almost incomprehensible.
I used to quite like SOAP, but after trying to actually write stuff using it, I've decided it's (to borrow a phrase i saw in a newsgroups) "a dog turd served on a fine china plate". The W3C produce some good specs, but this isn't one of them. I, of course, blame Microsoft:)
Oh and BTW, XML-RPC doesn't have xml as a first class data type either. Dunno why. I might write my own rpc protocol at this rate.
Pretty much correct, although quite a few GFX cards do in fact accelerate 2D operations like bezier curves and transparency.
For instance the Matrox Linux drivers will accelerate bezier curves in the near future, and can do accelerated alpha blending today. The problem is that these features are underutilized in most cards, for years it's been 3D games driving graphics technology forwards, so for instance the Matrox cards are the only ones that can do it (under Linux at any rate).
It's interesting that OSX is more useful as a desktop Unix than Linux is (for the non-technically-inclined user, someone who may be technically competent but not used to ripping things apart and making them work when they're broken) even though it's fairly new, whereas Linux has many years on it and still has a lot of stability, speed, compatibility, and usability problems as far as the desktop goes.
That's not really interesting. Every closed OS has this attribute with the exception of Windows, which has problems simply because it's so large and has been around for so long. Look at BeOS - they got it "just right" too, and it gained a following of semi-fanatical devotees. RISC OS just worked, as does Pocket Windows.
The reason Linux is still "getting there" with the desktop compared to MacOS is simply because Apple spent years with all the developers under one roof, working to one vision, with lots of money and talent thrown at it. Linux has been built by the hard work of volunteers who have mostly never met, where the freedom that comes with it means people do their own thing, and where even today I can count the number of professional artists that work on it on the fingers of my right hand! So comparing MacOS desktop development to Linux desktop development is apples and oranges.
So yes, since it aims to fulfill all our dreams of what an OS should be (fast (maybe), easy (yes), powerful (certainly), stable (maybe)) it does justify this number of stories, and more. We have traditionally been informed every time a new linux kernel comes out, and MacOSX will directly touch more lives than linux will any time soon.
Hmm, can't agree with that. BeOS was all those things, including fast and stable, and also ran on lots of hardware etc etc. They bombed of course, they were only a little company going against the inertia of an entire industry. Apple are larger, have pots of cash, and a fanatically loyal userbase, and even they are struggling (look at their financial reports). OSX is just BeOS all over again, but on a larger scale. I don't think Apple will go bust, they have too much money left over from the early days, but I think it'll be as important in the long term.
By the way, you forgot one fairly critical os "dream" that quite a lot of us have - where is the openness, the freedom in that list? The real difference to me at any rate is the everybody-is-equal aspect of it, I can theme Windows XP and install Cygwin to get UNIX, I prefer WinAmp to iTunes anyway....... it's all just details compared to the underlying social structures which are far more important.
Mozilla isn't actually bloatware, people who attribute interface latency problems to the fact that it can include an IRC client are misunderstanding the problems.
For reference, on my box here at work in Windows XP Mozilla feels as responsive as any other app, in fact it uses the theming APIs too so it even looks like native apps (that makes more of a difference than you might think). In Redhat 8 at work (what i now use mostly) it also feels rather snappy, at least as fast as on Windows. On Linux at home it feels much slower, even though they're the same build, and on OS X I've found it feels even slower than that (getting to the unusably slow point). BTW, for that comparison I used the RadialMenus extension - that is a good way of getting a feel for how fast Gecko is throwing boxes around the screen.
The slow UI speed of Mozilla on OS X is caused almost entirely by lack of optimizations in the low level Gecko Quartz drawing layer. Linux had similar problems in the 0.9.x releases, Mozilla was seriously abusing X (a sample trace i saw suggested it was redrawing parts of the screen several times over). Those kind of issues have been largely resolved lately, in part because companies like RedHat have been paying people to work full time on good Mozilla support, and partly because the for the longest time Mozilla Linux builds were shipping without even -O2 optimizations (a gcc flag) because it caused instability on a small number of machines. I think that whatever those instability issues were, they've been resolved, as Mozilla on Linux is now acceptably fast.
Remember that Mozilla is mostly load-on-demand, ie stuff like the mail components, the irc components (even the xul components) are loaded only when needed. People who accuse it of bloatware tend to overlook the fact that the only components Navigator loads that say Chimera doesn't are the XUL objects, and the RDF template engine (a part of xul) - combined these start in easily under a second, and don't really affect performance once underway as Gecko is very fast.
Apples response to this issue could have been, "we'll make Mozilla faster on OS X", but instead it was "we'll write a native front end to it". That's a lot of duplicated code: for what? Lickable widgets that Mozilla had support for anyway? Developing Chimera was the popular option, but looking with a wide perspective they could have got better results by sponsoring the aqua native widgets effort and improving the Mozilla quartz layer.
Linux has one thing that is totally ubiquitous and works, where the Windows equivalent is not. It's called X and would work wonders for this kind of thing [*]
Let's say you have a wireless network, and your "computer" is really simply a thin client that is driven remotely via X from a "desktop server". There are lots of compelling reasons for this setup anyway in corporate networks, but it's ideally suited for tablet PCs.
Imagine on your desk there are two towers, which hold a detachable flat panel. This panel really is flat too, and light. It contains only the display circuitry and a small, low powered chip with some software in flash ROM. These things exist today in the form of the PDA but I'm imagining tablet size here. The tablet/panel runs only a miniture X server. When docked to your thin client on your desk, the X server detects that there's also a keyboard near by and uses that. When there is no keyboard, it starts relaying pen messages to the desktop server and back comes your handwriting. Because you're not lugging around an actual computer, they can be fast, light, small and have long battery life. Because all your doing is moving X displays around, you've still got access to all your applications, all your documents just as if you were at your desk.
What's more, with some smart use of xmove, you can "throw" applications to another tablet. If you're running say a mapping application (or any specialist) and want to take it to a meeting, you just detach the screen and walk down the corridor. When there, you can share the app with others so they too can draw on it, or you can throw it over the the projector etc.
And yes, X can deal with pen input very nicely, it's then just a case of hooking up some good handwriting recognition to it on the server side. I think that'd be too cool.
[*]: Yes, I know windows has terminal services, but that works in a different way iirc and isn't as efficient bandwidth wise. Also some apps have difficulty with it and you I don't think stuff like OpenGL works too well. Anyway, my point stands. It can compete at any rate.
it seems that many companies have tried OSS software (or at least costed it) and have come to the same conclusions--in the long run, it's at least as expensive as commercial equivalents.
False logic - they may have costed individual open source apps and found they were more expensive than particular proprietary apps in some circumstances, but the opposite is also true. Nobody in their right mind should make generalizations like "we've costed open source software and it's more expensive than proprietary software"
# One (the one that oss zealots will jump on), yes it can be more expensive because of switching costs. that's only a small part of it though.
You're quite right they'll jump on it, because it's not really an argument at all. You said it yourself, not only is it small but it's also shrinking all the time as compatability gets better etc.
Two, it can be measurably (in a taylorist sort of way) more expensive to use OSS desktop applications because they are not designed with anywhere nearly the usability in mind that commercial aps are (note: A GUI != usability). I mean, if it takes my employees 10 minutes more a day to do their tasks with StarOffice or whathave you, then the cost of Ms-Office is soon worth it.
Usability is todays problem. Yesterday it was lack of desktop applications, and the day before that it was lack of corporate credability in the server arena. This isn't an argument against open source per se, it's merely a rather subjective statement about the state of some open source apps today. It's also another broad generalization, I find Redhat/GNOME2 to be more usually more intuitive than Windows. There are a lot of apps that have poor usability, but then the same can be said for commercial software (shareware anybody?)
Three, because of relatively poor usability of OSS development tools (whatever you may say, there are few OSS development environments that can come close to the better codewarrior or visual studio stuff), it is often more cost effective to develop in-house software on commercial platforms
Troll I say! What development tools you use are totally personal, I find Emacs/PyGNOME to be make me far more productive at desktop apps than IDEs such as Delphi or VS.NET. If you think that, then you're looking at the lack of wizards and "enterprise support" and assuming different is the same as inferior. If IDEs work for you then great, you can use Eclipse or Kylix on Linux if you must have an uber-powerful IDE.
remember the old saying.... "It's only free software if your time is not worth anything."
That's not an old saying, that's merely more FUD. It's also extremely arrogant, I use free software for everything and get paid to use it. Was I more productive when I used proprietary stuff to do my job? No. And my time is most definately worth something.
SAML is a protocol/framework for exchanging security assertions. It's not possible to build Passport out of pure SAML, for one SAML lacks a single signout protocol which kind of makes the whole thing rather useless. The Liberty Alliance (who will be releasing 1.1 soon) extend SAML to bring it up to speed.
We can basically forget about Passport interop for now. I did look into it a few weeks ago for the Identity system I'm working on, but unless Microsoft radically change things (and indications are they won't) anything more advanced than automatic logins would require their approval, you'd probably just get denied access to the network.
Re:Aren't APPS the real issue?
on
Halloween VII
·
· Score: 1
You seem to be under the impression that "Linux" is doing one thing. Not so! If you happen to like the MS way of doing things, or want to minimize retraining costs etc, you can:
Use KDE with OpenOffice and Konqueror like this......
... or use GNOME2, with AbiWord/Gnumeric like this (a bit less windowsish)
...... or use something totally far out like this. You have a truckload of flexibility as far as UIs are concerned.
The best software on *nix does it differently. Look at Apache. Anyone who wants it can figure out how to edit an httpd.conf file. It's not terribly hard. Why would anyone want to give it an IIS-like interface?
Because some people prefer GUIs? GUIs aren't evil you know, and editing text files aren't necessarily inferior either.
As an aside, anyone notice how much better the command-line paradigm deals with chaining/piping programs together?
No. The command line sucks at program componentization. Piping is an incredibly crude component system - you can't even get the exit code of a program half way along a pipe without pain. Stuff like COM/CORBA/.NET is soooo much better. Note that not every command line app can be scripted/joined easily either, they have to be specifically designed for that.
Re:Are you kidding?
on
Halloween VII
·
· Score: 3, Insightful
Now, the majority of you will go you "you could've fixed this and that by editing this and changing that and rewriting these 30 lines of code yatta yatta yatta" but who gives a fuck. It's a computer. It's supposed to make my life A: Easier or B: More Fun.
Nah, I'm not going to say that. I don't really know what I'd say, all that stuff you mentioned (except for debian) has always worked fine for me and my friends. I had to go download the official nVidia drivers to get good performance, but you have to do the same on Windows (except from CD instead of the net). I'd guess sometimes Win2K installs go awry too - or do you really think Windows is perfection itself?
Until the day when Linux supports any and every piece of hardware the day it's released, the installations can be done my grandma, and all software hits store shelves in both Windows and Linux version on the same day, Linux will never be a contender for the desktop market.
Wrong. Linux already is a contender for the desktop market. Most people don't buy hardware on the day it's released, and for most apps there are either Linux equivalents, the maker is considering porting it or it's close to working under Wine. Note: I said *most*, not all. Most apps is good enough for most people.
Anyway, the next battle isn't for people like you, it's for the business desktop, where stuff like the cumulative cost and TCO of Windows (2k or XP) is a big issue. Maybe in a few years we'll have all those issues nailed, and then you'll find that Linux makes your computer easier and more fun too.
Would you please make un-annotated versions available?
No. As it is, my defense against a copyright-violation suit by Microsoft would have to make rather creative use of the exemptions in copyright case law relating to journalism, satire and commentary. I fear that making un-annotated copies available would place me at significant legal risk.
Hey ESR, can you provide the next doc with XML and a couple of stylesheets so I can eliminate your comments? Sometimes they are good, but mostly they are just irritating cheerleading - we are quite capable of drawing our own conclusions thanks.
Re:Are you kidding?
on
Halloween VII
·
· Score: 3, Insightful
A company with 40 billion in cash and a 20 billion/year business doesn't write 2 page strategy documents.
Do you want to bet? It wouldn't surprise me if they have a LOT of people working on knowing everything there is to know about Linux, it's only smart considering that it's the only real competition.
This document is nothing special, in fact it's so boring I'd be surprised if it was faked. Maybe it was leaked intentionaly, maybe not. I dunno. What I do know is that it's quite likely they do multi-national surveys on Linux awareness/favourability quite regularly. In fact, one thing that puzzles me is that the tone of the document is one of "we haven't done this before", and this line:
"The study fielded between late-July and September 2001"
makes me think huh?? Who the hell waits a year before presenting a 2 page document like this? It can't possibly have taken a year to analyse a survey like that.
Anyway, don't have any doubt that they know all about Linux. They have some people monitoring the wine-devel lists as it is, and apparently a large number of their senior tech guys are well briefed about it. I know one of the Passport heads is familiar with WineX:) I expect they have people who's job actually involves reading slashdot. Funny in a way, but why wouldn't they? Easiest place to get news on open source happenings.
/me waves - hi there Microsoft person! Hope you're having fun:)
And what are people like me supposed to do? I live in a tiny flat and my PC basically IS my entertainment system. It plays DVDs, games, let's me work, acts as my radio (live next to a flipping great big hill so I have to use it as my radio) and plays my music. Assuming that anybody who sticks a CD into a computer is going to pirate it is ludicrous. I guess they'd expect me to buy a "real" CD Player or something dumb.
This is just getting more and more stupid. I'm not going to go download stuff from Kazaa just get, for one the effort it'd take to get it going in Wine combined with the general nastyness of the software and illegallity of it has put me off until now. I'm waiting for (and soon hopefully doing something about) the gift economy as a new model for music distribution, but there are quite a few technical and social hurdles to overcome first.
How long can the music industry keep this up though before what happened to Microsoft with Linux happens to the RIAA - the little people come out of the woodwork and come up with something new? Not long at this rate. Not long at all.
There was a time when I actually liked slashdot. Lately however, it's become so Anti-Microsoft - it's almost ridiculous. Keep in mind that there are over 60 thousand employees at Microsoft. All of us have families and mortgages to pay just like you.
My heart bleeds. Perhaps Slashdot has become so anti-microsoft lately because Microsoft have been ramping up the rate at which they give people reasons to dislike them?
Now of course, we could assume that as Mozilla becomes stabler and stabler, the filings should now slow down logarithmically, making the filing so late that we'll have have switched to Phoenix 4.0+gno/kMutt in the meantime...
Actually rate of bug filings speeds up as Mozilla gets more stable. It seems counterintuitive, but as Mozilla gets better more people use it, and so you get a) more dupes, b) more feature requests and c) preexisting bugs are found faster and more times.
You might have to adjust your equations slightly:)
Slashdot Mirror
Sigh. And now it's at -1.
I think it's fascinating to watch kneejerk moderators at work: I don't understand why certain people get so fanatical about products (and this applies to many products, not just computers) so I ask if anybody else does and get modded as flamebait. Yes, it's clearly flamebait, as in an Apple topic anything that doesn't praise his Jobsness is obviously inflammatory.
Jesus guys, that's my first -1 for over 8 months now, I do not post flamebait. I might have guessed my next would come from Apple fanboy moderators.
Actually, minority OS users do own the net :) Try doing a google search (and this works on altavista too) for Wine - the first link goes to WineHQ for the Windows emulator.
Now please don't give me any BS about how it's because the Mac is "so high quality" and "a classy piece of gear" as a Machead once tried to convince me, you can get high quality cars, stereo systems, hell even high quality food, but you don't see people build shrines to the Porsche do you? Or maybe you do.....
If anything I can kind of understand Linux freaks, as it's more a movement (vive la revolution and all that) than a product. But the Mac isn't anything other than some heavily marketed electronics. What motivates these people?
Huh? JavaScript is the Mozilla implementation of ECMAScript, a standard (not W3C) invented by Netscape. The DOM was also a Netscape idea, now standardized. CSS was originally proposed and largely designed by a guy from Opera. There are quite a few implementations out there actually, the idea that W3C technologies are too large to implement is crazy. Look at Mozilla, Amaya, even Konqueror is getting there now.....
The W3C should have stopped with a full specification of HTML. Anything they have been doing beyond that has been doing more harm than good. The web succeeded because HTML was simple.
Yes, and now it's ubiquitous do you really think we need to keep it simple? Being simple was great when the web was small, it let it grow very quickly. Why should we keep it simple now? Just for the sake of it? I'd rather have power. If that means there are only 3 or 4 quality implementations as opposed to 20, then so be it.
The world is not a simple place, and the things we want to do with the web nowadays aren't simple either. If you want simplicity then feel free to write a web browser that only understands a subset of the standards, they are layered so people can do this. Just bear in mind that it won't be useful for browsing the web, because a lot of people like powerful technologies and use them.
You have to have standards. The W3C are the people who are widely recognized as being the technical lead for the net. Now they don't make law, quite right, but if there was no W3C then Microsoft really WOULD own the web: as it is, we can and do take them to task when they break the rules. They can ignore us of course, yet whaddaya know but IE6 supports DOM/CSS Level 1. Not a particularly impressive achievement, but it's a start.
The standards are actually very precise, which is one reason they are seens as being very large. There is hardly any room for interpretation in stuff like the DOM, CSS, XML etc. Of course, sometimes when the internal architecture of IE mandates it Microsoft simply ignore things, the mime-type issue being a good example, but also the fact that you have to specify node.className = "class" to set the style on a new element, as opposed to setting the class attribute (which works fine in Mozilla). Why? Because (according to an MS developer) internally the MS dom is based on object model attributes, so that's what you have to set.
Has any company yet written a complete CSS1 implementation? A complete working version of DOM0? Yet here we are toiling away on XHTML and CSS3(!) and DOM Level 2. And they don't even seem to give a rat's ass if anyone actually follows the rules.
[sigh] Yes. Mozilla supports DOM and CSS Level 2 and they have partial support for Level 3 now. Level 0 is the term used to refer to the pre-standardized technologies, it doesn't actually exist as a standard so EVERY browser that can script web pages has a level zero DOM. It should be noted that TBL himself has stepped in on occasion to tick off Microsoft about stuff like browser blocks, bad HTML etc.
From what I hear about CSS3, it's going to be such a massive specification that no company (save Microsoft, if they actually gave a damn) would possibly be able to implement it.
Then you hear wrong.
In the meantime we see developers actually building websites entirely out of Flash because there's one reference implementation (one version, period) and it just works. Is that the future we want?
Developers do not build web pages out of flash. Marketing departments do. Luckily most web pages are not built by marketing.
It's time to hold these clowns accountable. Make them do some real work: make them create a working version of their spec.
Poor troll. The W3C already implement all their standards, go to w3.org and download Amaya. Nobody uses it for actually browsing the web, but there it is, proof that an actually very small organization with very few coders can implement their standards.
This whole situation has been caused simply because a network stream is a 1:1 thing. If you make it a 1:many connection then suddenly you have far more efficient use of bandwidth, because if you're an ISP and 100 users are downloading a file, you only have to receive it once on your incoming link, instead of 100 times.
Of course, you'd still have P2P file sharing, the reason that people use Kazaa is not because it's the best way of moving information around (it's not), it's because it's anonymous and you can't be taken down for it. Safety in numbers, safety in anonymity. If there were suddenly large pirate music servers transmitting albums on rotation via multicast 24/7 they would be much easier targets.
Multicast has lots of other legal uses of course, that's what I'd want them for. But I can see that it'd help solve this situation. So come on ISPs, where are the v6 routers?
The simplest solution is simply to place upload caps. You wouldn't eliminate all p2p users, but that's OK, you just want to hit the guys who constantly upload stuff 24/7.
Unlikely. As it's already been pointed out above, if MS was to be sued under this patent, they'd simply pull ActiveX from IE (if it held, which it wouldn't). The world would continue to go around, albiet the girl in my office would wonder why she can't play scrabble on MSN Games anymore.
If Microsoft got these patents and sued Mozilla, what would happen? Jack all is my answer. Who would they sue? The Mozilla organization? Can they even be held legally responsible for Mozilla? And if they could, they still couldn't destroy Mozilla, it's an open source project, we all have it, and it would continue to be worked on.
Years ago MS was talking about using patents as a weapon against free software - but I've not seen anything yet. Have you?
You are defining ubiquitous how?
Ubiquitous within Linux, what did you think? I'm assuming a Linux based network.
Oh, and when will X get audio support? I can play back movies with sound through Terminal Server, even over my DSL. So maybe I don't have as great an experience with D3D via TS, but I've never been able to play Quake over X over DSL on X.
Huh? You're confused. X is a graphics protocol, audio is outside its remit. If you want network transparent audio you can setup aRts or eSound. It'd be nice to have it integrated with the DISPLAY variable or something, but that's pretty trivial to do.
As for Quake, I have no idea how you managed to play a fullscreen 3D game over terminal services (which is based on compressed bitmaps) on DSL, it must have been a very low latency line. You can play games over X though, I do know that, as it sends the opengl calls themselves through the connection.
On a related note, where are all those great applications for X on Linux? Things like Adobe Photoshop or Paint Shop Pro or Avid or Finale or IE. Can you name me one app for Linux that doesn't have a port or equivalent on Windows?
Tip: when trolling, don't pick bad examples. Adobe Photoshop/PSP = The gimp. Yes, that exists on Windows too, because somebody ported it. That's software freedom working for you. IE = Mozilla. Avid?? dunno what that is, ditto for Finale. There are almost certainly equivalents for joe-level apps, if there are specialists things you need you can use Wine. How did we get onto this anyway, is this just some kind of general anti-linux rant? I was talking about applying X to tablet pcs.
Your thumbnails example is a terrible one. You're Linux geek friend was probably in the middle of doing something wierd to his machine, if I wanted to view a CD of images (i'm assuming they're in a standard format, not one MS pulled out of its ass) then I just put the CD in the drive and Nautilus or Konquerer will open and compile thumbnails for me. I know from experience they do this very quickly indeed. Maybe your friend eschews the easy way or something, I dunno.
I really have to laugh at anyone who thinks they can be better than MS at integrating technologies around Linux and competing on the Tablet platform before Microsoft eats it all.
Laugh then. We'll laugh last, guaranteed. This is such a pathetic pro-MS rant it's untrue. Why am I bothering to reply? Oh, I know, it's raining and I don't want to get wet by going home just yet.
If Microsoft has never, ever let you down then by definition you are new to the computing scene. Having friends who can get you cheap copies of Windows is pathetic, and that last comment is ludicrous - I got redhat for free! And I didn't break the law or call favours. It's more stable than Windows, virus free, lots of good support etc etc etc.
Jesus. I give up. I hope you weren't serious, if so then you should really do some investigations instead of trolling slashdot, you might learn something.
Live queries - coming, at least in ReiserFS and apparently Longhorn. I've not used BeOS but I find simply organizing my files well into directories means I never want to do searches anyway.
Version control in filing systems isn't a new idea, but again, the devil is in the details. What happens if you run out of disk space (i for instance have only a 10gig disk and it's already full). Do you start randomly killing past revisions? How do you store the different versions? Plain text is easy, you just diff it, but what about images? Store a whole new copy of the image each time perhaps, because as far as I know there is no image diffing technology.
Perhaps a better idea is to store transactions. Any program that supports multi-level undo probably internally is structured around transactions, ie objects that represent an alteration to the document. A document could be split up into 2 parts: the data itself and a transactions queue. When you start a new file, the file stores your actions, rather than the actual document state, and loading it involves rebuilding the document from the transactions. This would allow theoretically infinite undo, limited only by disk space. Of course rerunning actions would take ages, so you could have a throwaway cache of the current state of the document.
Erm, what else? The RISC OS style filepickers are available for use in your apps via the rox library, and yes, they do rock. You can do this in Windows at any rate by opening up an explorer window, finding the document and dragging it onto the window. It doesn't work very well however. It would be relatively easy to add XDS support to the GTK file picker (it's being rewritten by hadess i think) so if you want you can use drag and drop save/load.
Application loading/unloading - there's no real reason why we should keep this, he's quite right, and really we don't have it anymore. When I want to browse the web, I click on the web browser button in my panel, and up pops a Mozilla window. If Moz isn't loaded it takes a few more seconds than normal, but so what? It's multi-instance. When I logout, my apps persist themselves to the session manager so they are back again when I login. I don't think much about loading/unloading apps anymore, that's just a dumb Mac thing - one they could kill so easily, but didn't.
What MPT is really ranting about here is not what he thinks he is. Computers have history, and with history comes inertia. Support linux man! You can add most of what you've been talking about to apps pretty easily when all the code is open. I can think of ways of doing all those things with slight modifications to toolkits and apps.
When a frustrated person on the list asked the same questions I'd been having, he was told there seemed to be an informal standard of using an Apache Axis namespace for XML now! I mean WTF? But that's only the beginning.
Where is objects-by-reference? Oh, I know, that'd be in the Microsoft .NET Remoting extensions. Why not in the standard itself? Because MS have been blocking it on the grounds of "added bloat". I'd say objects-by-reference are absolutely critical to any modern RPC protocol and it doesn't have to be complex either, but it's not in there. Simple Object Access Protocol my backside.
And then of course the patents. Considering we've been doing RPC for over a decade, and SOAP is merely RPC done in XML (and done worse) how in the lords name can there be patents on it? Prior art trips you up at the street corner there's so much. And WSDL? What a POS. That is a classic example of something that's been abstracted so much it's almost incomprehensible.
I used to quite like SOAP, but after trying to actually write stuff using it, I've decided it's (to borrow a phrase i saw in a newsgroups) "a dog turd served on a fine china plate". The W3C produce some good specs, but this isn't one of them. I, of course, blame Microsoft :)
Oh and BTW, XML-RPC doesn't have xml as a first class data type either. Dunno why. I might write my own rpc protocol at this rate.
For instance the Matrox Linux drivers will accelerate bezier curves in the near future, and can do accelerated alpha blending today. The problem is that these features are underutilized in most cards, for years it's been 3D games driving graphics technology forwards, so for instance the Matrox cards are the only ones that can do it (under Linux at any rate).
That's not really interesting. Every closed OS has this attribute with the exception of Windows, which has problems simply because it's so large and has been around for so long. Look at BeOS - they got it "just right" too, and it gained a following of semi-fanatical devotees. RISC OS just worked, as does Pocket Windows.
The reason Linux is still "getting there" with the desktop compared to MacOS is simply because Apple spent years with all the developers under one roof, working to one vision, with lots of money and talent thrown at it. Linux has been built by the hard work of volunteers who have mostly never met, where the freedom that comes with it means people do their own thing, and where even today I can count the number of professional artists that work on it on the fingers of my right hand! So comparing MacOS desktop development to Linux desktop development is apples and oranges.
So yes, since it aims to fulfill all our dreams of what an OS should be (fast (maybe), easy (yes), powerful (certainly), stable (maybe)) it does justify this number of stories, and more. We have traditionally been informed every time a new linux kernel comes out, and MacOSX will directly touch more lives than linux will any time soon.
Hmm, can't agree with that. BeOS was all those things, including fast and stable, and also ran on lots of hardware etc etc. They bombed of course, they were only a little company going against the inertia of an entire industry. Apple are larger, have pots of cash, and a fanatically loyal userbase, and even they are struggling (look at their financial reports). OSX is just BeOS all over again, but on a larger scale. I don't think Apple will go bust, they have too much money left over from the early days, but I think it'll be as important in the long term.
By the way, you forgot one fairly critical os "dream" that quite a lot of us have - where is the openness, the freedom in that list? The real difference to me at any rate is the everybody-is-equal aspect of it, I can theme Windows XP and install Cygwin to get UNIX, I prefer WinAmp to iTunes anyway ....... it's all just details compared to the underlying social structures which are far more important.
For reference, on my box here at work in Windows XP Mozilla feels as responsive as any other app, in fact it uses the theming APIs too so it even looks like native apps (that makes more of a difference than you might think). In Redhat 8 at work (what i now use mostly) it also feels rather snappy, at least as fast as on Windows. On Linux at home it feels much slower, even though they're the same build, and on OS X I've found it feels even slower than that (getting to the unusably slow point). BTW, for that comparison I used the RadialMenus extension - that is a good way of getting a feel for how fast Gecko is throwing boxes around the screen.
The slow UI speed of Mozilla on OS X is caused almost entirely by lack of optimizations in the low level Gecko Quartz drawing layer. Linux had similar problems in the 0.9.x releases, Mozilla was seriously abusing X (a sample trace i saw suggested it was redrawing parts of the screen several times over). Those kind of issues have been largely resolved lately, in part because companies like RedHat have been paying people to work full time on good Mozilla support, and partly because the for the longest time Mozilla Linux builds were shipping without even -O2 optimizations (a gcc flag) because it caused instability on a small number of machines. I think that whatever those instability issues were, they've been resolved, as Mozilla on Linux is now acceptably fast.
Remember that Mozilla is mostly load-on-demand, ie stuff like the mail components, the irc components (even the xul components) are loaded only when needed. People who accuse it of bloatware tend to overlook the fact that the only components Navigator loads that say Chimera doesn't are the XUL objects, and the RDF template engine (a part of xul) - combined these start in easily under a second, and don't really affect performance once underway as Gecko is very fast.
Apples response to this issue could have been, "we'll make Mozilla faster on OS X", but instead it was "we'll write a native front end to it". That's a lot of duplicated code: for what? Lickable widgets that Mozilla had support for anyway? Developing Chimera was the popular option, but looking with a wide perspective they could have got better results by sponsoring the aqua native widgets effort and improving the Mozilla quartz layer.
Linux has one thing that is totally ubiquitous and works, where the Windows equivalent is not. It's called X and would work wonders for this kind of thing [*]
Let's say you have a wireless network, and your "computer" is really simply a thin client that is driven remotely via X from a "desktop server". There are lots of compelling reasons for this setup anyway in corporate networks, but it's ideally suited for tablet PCs.
Imagine on your desk there are two towers, which hold a detachable flat panel. This panel really is flat too, and light. It contains only the display circuitry and a small, low powered chip with some software in flash ROM. These things exist today in the form of the PDA but I'm imagining tablet size here. The tablet/panel runs only a miniture X server. When docked to your thin client on your desk, the X server detects that there's also a keyboard near by and uses that. When there is no keyboard, it starts relaying pen messages to the desktop server and back comes your handwriting. Because you're not lugging around an actual computer, they can be fast, light, small and have long battery life. Because all your doing is moving X displays around, you've still got access to all your applications, all your documents just as if you were at your desk.
What's more, with some smart use of xmove, you can "throw" applications to another tablet. If you're running say a mapping application (or any specialist) and want to take it to a meeting, you just detach the screen and walk down the corridor. When there, you can share the app with others so they too can draw on it, or you can throw it over the the projector etc.
And yes, X can deal with pen input very nicely, it's then just a case of hooking up some good handwriting recognition to it on the server side. I think that'd be too cool.
[*]: Yes, I know windows has terminal services, but that works in a different way iirc and isn't as efficient bandwidth wise. Also some apps have difficulty with it and you I don't think stuff like OpenGL works too well. Anyway, my point stands. It can compete at any rate.
False logic - they may have costed individual open source apps and found they were more expensive than particular proprietary apps in some circumstances, but the opposite is also true. Nobody in their right mind should make generalizations like "we've costed open source software and it's more expensive than proprietary software"
# One (the one that oss zealots will jump on), yes it can be more expensive because of switching costs. that's only a small part of it though.
You're quite right they'll jump on it, because it's not really an argument at all. You said it yourself, not only is it small but it's also shrinking all the time as compatability gets better etc.
Two, it can be measurably (in a taylorist sort of way) more expensive to use OSS desktop applications because they are not designed with anywhere nearly the usability in mind that commercial aps are (note: A GUI != usability). I mean, if it takes my employees 10 minutes more a day to do their tasks with StarOffice or whathave you, then the cost of Ms-Office is soon worth it.
Usability is todays problem. Yesterday it was lack of desktop applications, and the day before that it was lack of corporate credability in the server arena. This isn't an argument against open source per se, it's merely a rather subjective statement about the state of some open source apps today. It's also another broad generalization, I find Redhat/GNOME2 to be more usually more intuitive than Windows. There are a lot of apps that have poor usability, but then the same can be said for commercial software (shareware anybody?)
Three, because of relatively poor usability of OSS development tools (whatever you may say, there are few OSS development environments that can come close to the better codewarrior or visual studio stuff), it is often more cost effective to develop in-house software on commercial platforms
Troll I say! What development tools you use are totally personal, I find Emacs/PyGNOME to be make me far more productive at desktop apps than IDEs such as Delphi or VS.NET. If you think that, then you're looking at the lack of wizards and "enterprise support" and assuming different is the same as inferior. If IDEs work for you then great, you can use Eclipse or Kylix on Linux if you must have an uber-powerful IDE.
remember the old saying.... "It's only free software if your time is not worth anything."
That's not an old saying, that's merely more FUD. It's also extremely arrogant, I use free software for everything and get paid to use it. Was I more productive when I used proprietary stuff to do my job? No. And my time is most definately worth something.
Passport is a centralized web based SSO system.
SAML is a protocol/framework for exchanging security assertions. It's not possible to build Passport out of pure SAML, for one SAML lacks a single signout protocol which kind of makes the whole thing rather useless. The Liberty Alliance (who will be releasing 1.1 soon) extend SAML to bring it up to speed.
We can basically forget about Passport interop for now. I did look into it a few weeks ago for the Identity system I'm working on, but unless Microsoft radically change things (and indications are they won't) anything more advanced than automatic logins would require their approval, you'd probably just get denied access to the network.
Use KDE with OpenOffice and Konqueror like this ......
The best software on *nix does it differently. Look at Apache. Anyone who wants it can figure out how to edit an httpd.conf file. It's not terribly hard. Why would anyone want to give it an IIS-like interface?
Because some people prefer GUIs? GUIs aren't evil you know, and editing text files aren't necessarily inferior either.
As an aside, anyone notice how much better the command-line paradigm deals with chaining/piping programs together?
No. The command line sucks at program componentization. Piping is an incredibly crude component system - you can't even get the exit code of a program half way along a pipe without pain. Stuff like COM/CORBA/.NET is soooo much better. Note that not every command line app can be scripted/joined easily either, they have to be specifically designed for that.
Nah, I'm not going to say that. I don't really know what I'd say, all that stuff you mentioned (except for debian) has always worked fine for me and my friends. I had to go download the official nVidia drivers to get good performance, but you have to do the same on Windows (except from CD instead of the net). I'd guess sometimes Win2K installs go awry too - or do you really think Windows is perfection itself?
Until the day when Linux supports any and every piece of hardware the day it's released, the installations can be done my grandma, and all software hits store shelves in both Windows and Linux version on the same day, Linux will never be a contender for the desktop market.
Wrong. Linux already is a contender for the desktop market. Most people don't buy hardware on the day it's released, and for most apps there are either Linux equivalents, the maker is considering porting it or it's close to working under Wine. Note: I said *most*, not all. Most apps is good enough for most people.
Anyway, the next battle isn't for people like you, it's for the business desktop, where stuff like the cumulative cost and TCO of Windows (2k or XP) is a big issue. Maybe in a few years we'll have all those issues nailed, and then you'll find that Linux makes your computer easier and more fun too.
Would you please make un-annotated versions available?
No. As it is, my defense against a copyright-violation suit by Microsoft would have to make rather creative use of the exemptions in copyright case law relating to journalism, satire and commentary. I fear that making un-annotated copies available would place me at significant legal risk.
Hey ESR, can you provide the next doc with XML and a couple of stylesheets so I can eliminate your comments? Sometimes they are good, but mostly they are just irritating cheerleading - we are quite capable of drawing our own conclusions thanks.
Do you want to bet? It wouldn't surprise me if they have a LOT of people working on knowing everything there is to know about Linux, it's only smart considering that it's the only real competition.
This document is nothing special, in fact it's so boring I'd be surprised if it was faked. Maybe it was leaked intentionaly, maybe not. I dunno. What I do know is that it's quite likely they do multi-national surveys on Linux awareness/favourability quite regularly. In fact, one thing that puzzles me is that the tone of the document is one of "we haven't done this before", and this line:
"The study fielded between late-July and September 2001"
makes me think huh?? Who the hell waits a year before presenting a 2 page document like this? It can't possibly have taken a year to analyse a survey like that.
Anyway, don't have any doubt that they know all about Linux. They have some people monitoring the wine-devel lists as it is, and apparently a large number of their senior tech guys are well briefed about it. I know one of the Passport heads is familiar with WineX :) I expect they have people who's job actually involves reading slashdot. Funny in a way, but why wouldn't they? Easiest place to get news on open source happenings.
This is just getting more and more stupid. I'm not going to go download stuff from Kazaa just get, for one the effort it'd take to get it going in Wine combined with the general nastyness of the software and illegallity of it has put me off until now. I'm waiting for (and soon hopefully doing something about) the gift economy as a new model for music distribution, but there are quite a few technical and social hurdles to overcome first.
How long can the music industry keep this up though before what happened to Microsoft with Linux happens to the RIAA - the little people come out of the woodwork and come up with something new? Not long at this rate. Not long at all.
My heart bleeds. Perhaps Slashdot has become so anti-microsoft lately because Microsoft have been ramping up the rate at which they give people reasons to dislike them?
Actually rate of bug filings speeds up as Mozilla gets more stable. It seems counterintuitive, but as Mozilla gets better more people use it, and so you get a) more dupes, b) more feature requests and c) preexisting bugs are found faster and more times.
You might have to adjust your equations slightly :)